* Refactor to split `ChainSync::sync` method in two
Replace the use of loop labels and `continue` for control flow, and use
early return from a separate method instead. This also allows removing
the `started_once` flag.
* Refactor to create `handle_block_response` helper
Reduce duplicate code and make the main synchronization methods a little
more concise to improve readability.
* Only cancel downloads in case of error
Leave active downloads running if the tips have been exhausted, because
it could have reached the chain tip.
* Add the ChainTip and Network to RpcImpl
* Add the new RpcImpl fields to the tests
* Simplify RPC version field using generics
* Temporarily allow unused struct fields
* Use the read-only state service for RPCs
* Refactor non-finalized block lookup into Chain
* Implement the read-only state block request
* Drop the Chain watch channel lock before accessing the finalized state
* feat(rpc): implement `getrawmempool` rpc method
* tests(rpc): start implementing a test
* Return a `Vec<String>` from `get_raw_mempool`
Ensure the returned type is serialized as an array of hexadecimal
strings.
* Make `getrawmempool` test a property test
Test it with random sets of transaction IDs.
* Fix code formatting inside `proptest!`
Run `rustfmt` inside the `proptest!` block.
* tests(rpc): fix lightwalletd test
* fix(rpc): remove non used return type
* tests(rpc): remove assert
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Allow extra time for lightwalletd integration tests
* Wait for a successful block ingestion in lightwalletd integration
* Check for any ingestor block, not just block 1
* Add state service module docs and cleanup
* Move and add finalized state methods
* Add chain and non-finalized state methods
* Cleanup methods and imports
* Create a ReadStateService type
* Add a stub service implementation
* Add a TODO
* Update ReadStateService request stubs with RPC names and tickets
* Documentation updates
* Make RPC State generic bounds accept a buffered state and a read-only state
* Doc updates
* Add missing proptest-impl feature in RPC dev dependencies
* feature(rpc): start adding a `getblock` method
* fix(rpc): replace oneshot
* fix(rpc): replace a panic with error
* fix(rpc): fix test
* feature(rpc): add hex to response
* refactor(rpc): use generic instead of alias
* docs(rpc): improve docs for getblock method
* test(rpc): add a test for getblock method
* deps(rpc): remove non needed tower features
Co-authored-by: teor <teor@riseup.net>
* docs(rpc): add a note to getblock doc
* refactor(rpc): replace alias
* fix(rpc): use `zcash_serialize_to_vec()` instead of logging format
* tests(rpc): add network argument to `populated_state()`
* refactor(rpc): use an error for state service readiness
* fix(rpc): add parameter
* fix(rpc): clippy
* nit(rpc): remove new line from imports
* fix(rpc): remove commented code
* fix(rpc): simplify error
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Use a `SerializedBlock` type to help serializing blocks (#3725)
* Create a `SerializedBlock` helper type
Create a type that can be used as a byte slice, but is guaranteed to
represent a valid block.
* Use `into_iter` instead of `iter`
There's no need to borrow the elements, they can be moved out directly.
This will be necessary because `&Arc<T>` doesn't implement `Borrow<T>`,
so a `SerializedBlock` can't be built directly from an `&Arc<Block>`.
* Use `SerializedBlock` in `GetBlock`
Make the type stricter to avoid storing possibly invalid values. The
bytes are still serialized as a hexadecimal string, through the usage of
`hex`.
The `serde::Deserialize` can't be derived because `hex` requires the
type to also implement `FromHex`.
* feature(rpc): add suggestions from code review
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* tests(rpc): make sure mempool has no requests in get_block test
* fix(rpc): change height argument type in getblock method
* fix(rpc): rustfmt
* fix(rpc): replace panic
* fix(rpc): change getblock response
* fix(rpc): fix lightwalletd test
* tests(rpc): add a getblock error test
* fix(rpc): try another regex
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Stub `sendrawtransaction` RPC method
Register the RPC method, and stub an implementation that currently just
panics. The method has a single `String` parameter with the hexadecimal
string of the raw transaction's bytes and returns a
`SentTransactionHash` wrapper type that's just a hexadecimal `String` of
the sent transaction's hash.
* Add mempool service instance to `RpcImpl`
Use a type parameter to represent the mempool service using the
interface defined by `zebra-node-services`.
* Update test vector to use a mock mempool service
Update the test to be compatible with the changes to `RpcImpl`. The mock
mempool service is expected to not be used during the test.
* Use a `tower::Buffer` for the mempool service
Make it simpler to send requests to the service in a concurrent manner.
* Return a `Future` from `send_raw_transaction`
Make the call asynchronous.
* Implement `sendrawtransaction` RPC
Deserialize the transaction and send it to be queued for verification
and subsequent inclusion in the mempool.
* Test if mempool receives sent raw transaction
Use a mock service as the mempool service and check that it receives a
sent raw transaction.
* Test using non-hexadecimal string parameter
The method should return an error.
* Test with bytes that fail deserialization
Check that the method returns an invalid parameters error if the input
can't be deserialized as a `Transaction`.
* Test if mempool errors are forwarded to caller
Mempool service errors should be sent back to the remote caller as
server errors.
* Test transactions rejected by the mempool service
Transactions that are rejected by the mempool service should result in
a server error being sent to the caller.
* Improve error message
Add the word "structurally" to make it clear that the issue is in the
transaction's deserialization.
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
* Add note regarding missing `allowhighfees` param.
The parameter isn't supported yet because `lightwalletd` doesn't use it.
* Update the documentation to be consistent
Follow the convention adopted by the `get_info` RPC method.
* Implement `ToHex` and `FromHex` for `Hash`
Make it easier to generate hexadecimal strings from `transaction::Hash`
instances.
* Use `ToHex` in `Debug` and `Display`
Reduce repeated code.
* Refactor to add `bytes_in_display_order` method
Use it to remove repeated code and improve clarity a bit.
* Use `hex::serialize` to serialize transaction hash
Make the type stricter in its contents, while still serializing the
transaction has as a hexadecimal string.
* Simplify serialization attribute
Deserialization should also use `hex::deserialize`, so using the shorter
attribute makes things easier to read and more future proof.
* Update zebra-chain/src/transaction/hash.rs
* Remove unnecessary lifetime
The anonymous lifetime is automatically inferred by the compiler.
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
Sometimes lightwalletd goes straight to "Another refresh in progress",
without logging "Mempool refresh error" first.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* Upgrade some dependencies
* Upgrade some dependencies
* Upgrade dependencies for zebrad
* Upgrade tracing dependencies
* Revert `tor` & `arti`
* Upgrade `criterion` & `pin-project` in `deny.toml`
* Remove some dependencies from `skip-tree` in `deny.toml`
* Revert some the versions of dependencies because of duplicates
* Revert proptest regressions
* Upgrade dependencies, then ignore some more duplicates (#3716)
* feat(actions)!: add full sync test (#3582)
* add(tests): full sync test
* fix(test): add build
* fix(deploy): escape double dashes '--' correctly
* fix(test): remove unexpected --no-capture arg
error: Found argument '--nocapture' which wasn't expected, or isn't valid in this context
* refactor(docker): use default executable as entrypoint
* refactor(startup): add a custom entrypoint
* fix(test): add missing TEST_FULL_SYNC variable
* test(timeout): use the biggest machine
* fix
* fix(deploy): use latest successful image
* typo
* refactor(docker): generate config file at startup
* revert(build): changes were made to docker
* fix(docker): send variables correctly to the entrypoint
* test different conf file approach
* fix(env): add RUN_TEST env variable
* ref: use previous approach
* fix(color): use environment variable
* fix(resources): use our normal machine size
* fix(ci): double CPU and RAM for full sync test
* fix(test): check for zebrad test output in the correct order
The mempool is only activated once, so we must check for that log first.
After mempool activation, the stop regex is logged at least once.
(It might be logged before as well, but we can't rely on that.)
When checking that the mempool didn't activate,
wait for the `zebrad` command to exit,
then check the entire log.
* fix(ci): run full sync test with full compiler optimisations
* fix(tests): reintroduce tests and run full sync on approval
* fix(tests): reduce the changelog
Co-authored-by: teor <teor@riseup.net>
* fix(ci): update CI job path triggers (#3692)
* ci(test): re-run tests when snapshot data changes
* fix(ci): rebuild state when disk format changes
* fix(ci): rebuild rust docs when code or dependencies change
* doc(ci): explain why we run jobs when files change
Co-authored-by: Gustavo Valverde <gustavo@iterativo.do>
* fix(build): use the right multistage target (#3700)
* fix(review): only assign one reviewer to general Rust reviews (#3708)
If we assign two teams, GitHub assigns two reviewers.
* fix(ci): change the color-eyre ignore to a tracing-subscriber ignore
* fix(ci): ignore duplicate darling dependencies
* doc(ci): remove an alternative resolution doc
Co-authored-by: Gustavo Valverde <gustavo@iterativo.do>
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Gustavo Valverde <gustavo@iterativo.do>
* Stub `sendrawtransaction` RPC method
Register the RPC method, and stub an implementation that currently just
panics. The method has a single `String` parameter with the hexadecimal
string of the raw transaction's bytes and returns a
`SentTransactionHash` wrapper type that's just a hexadecimal `String` of
the sent transaction's hash.
* Add mempool service instance to `RpcImpl`
Use a type parameter to represent the mempool service using the
interface defined by `zebra-node-services`.
* Update test vector to use a mock mempool service
Update the test to be compatible with the changes to `RpcImpl`. The mock
mempool service is expected to not be used during the test.
* Use a `tower::Buffer` for the mempool service
Make it simpler to send requests to the service in a concurrent manner.
* Return a `Future` from `send_raw_transaction`
Make the call asynchronous.
* Implement `sendrawtransaction` RPC
Deserialize the transaction and send it to be queued for verification
and subsequent inclusion in the mempool.
* Test if mempool receives sent raw transaction
Use a mock service as the mempool service and check that it receives a
sent raw transaction.
* Test using non-hexadecimal string parameter
The method should return an error.
* Test with bytes that fail deserialization
Check that the method returns an invalid parameters error if the input
can't be deserialized as a `Transaction`.
* Test if mempool errors are forwarded to caller
Mempool service errors should be sent back to the remote caller as
server errors.
* Test transactions rejected by the mempool service
Transactions that are rejected by the mempool service should result in
a server error being sent to the caller.
* Improve error message
Add the word "structurally" to make it clear that the issue is in the
transaction's deserialization.
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
* Add note regarding missing `allowhighfees` param.
The parameter isn't supported yet because `lightwalletd` doesn't use it.
* Update the documentation to be consistent
Follow the convention adopted by the `get_info` RPC method.
* Remove mempool service usage line
It contained incomplete information that's not really necessary.
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Fix formatting
`rustfmt` was not executed on the file for the previous commit because I
had edited it on GitHub.
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* fix(ci): remove an unused trigger path
* doc(ci): explain lightwalletd trigger paths
* fix(test): check for adityapk00/lightwalletd behaviour in test harness
* fix(ci): work around buildx command error
* fix(ci): revert the workaround
* tests(rpc): add an rpc endpoint test
* tests(rpc): add an rpc port conflict test
* tests(rpc): refactor some imports
* tests(rpc): fix failures, make test more complete
* tests(rpc): parse json response for better coverage
* tests(rpc): change request
* tests(rpc): wait until port is open in rpc_endpoint test
* tests(rpc): add a delay between launching 2 nodes
* tests(rpc): try 5 seconds
* refactor(rpc): open rpc server faster
* tests(rpc): extend `LAUNCH_DELAY` to 15 seconds
* fix(rpc): disable rpc_conflict test for windows
* fix(ci): skip the RPC tests if the network is disabled
* rustfmt
* fix(zebrad/test): test function return type
* tests(rpc): print server output in assert
* fix(rpc): fix acceptance test looking for string in `build` field
* fix(rpc): reduce the number of acceptable characters in version output
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* fix(test): only run lightwalletd test when the ZEBRA_TEST_LIGHTWALLETD env var is set
* fix(test): actually skip the test
* doc(zebrad): add some test TODOs
* doc(test): document zebrad-specific process launch methods
* refactor(zebrad): split lightwalletd launch into its own testing methods
* fix(zebrad/test): simplify file writing
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* refactor(zebrad/test): rename argument variables
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* lint(zebrad/tests): remove unused import
* fix(zebrad/test): restore SocketAddr import that was removed on main
* rustfmt
* fix(zebrad/test): update integration test to match adityapk00/lightwalletd behaviour
* rustfmt
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* feature(rpc): add getinfo subversion field and getinfo docs
* feature(rpc): add getinfo build field
* refactor(rpc): replace the lazy_static
* docs(rpc): fic typo, add link to zcashd ticket
* tests(rpc): add getinfo unit test
* docs(rpc): complete comment
The mempool is only activated once, so we must check for that log first.
After mempool activation, the stop regex is logged at least once.
(It might be logged before as well, but we can't rely on that.)
When checking that the mempool didn't activate,
wait for the `zebrad` command to exit,
then check the entire log.
* Create new empty `zebra-node-services` crate
The goal is to store the mempool `Request` and `Response` types so that
the `zebra-rpc` crate can interface with the mempool service without
having to import `zebrad`.
* Move `Gossip` mempool type into new crate
It is required by the `Request` type, which will be moved next.
* Add documentation to `Gossip` variants
Avoid having to add an exception to allow undocumented code.
* Move `mempool::Request` type to new crate
The first part of the service interface definition. Usages have been
changed to refer to the new crate directly, and since this refactor is
still incomplete, some `mp` aliases are used in a few places to refer to
the old module.
* Create an `UnboxMempoolError` helper trait
Centralize some common code to extract and downcast boxed mempool
errors. The `mempool::Response` will need to contain a `BoxError`
instead of a `MempoolError` when it is moved to the
`zebra-node-services` crate, so this prepares the tests to be updated
with less changes.
* Use `UnboxMempoolError` in tests
Make the necessary changes so that the tests are ready to support a
`BoxError` in the `mempool::Response` type.
* Use `BoxError` in `mempool::Response`
Prepare it to be moved to the `zebra-node-services` crate.
* Move `mempool::Response` to `zebra-node-services`
Update usages to import from the new crate directly.
* Remove `mp` aliases for mempool component module
Use any internal types directly instead.
* Replace `tower::BoxService` with custom alias
Remove the dependency of `zebra-node-services` on `tower`.
* Move `Gossip` into a separate `sub-module`
Keep only the main `Request` and `Response` types in the `mempool`
module.
* Use `crate::BoxError` instead of `tower::BoxError`
Follow the existing convention.
* Add missing `gossip.rs` module file
It was missing from a previous refactor commit.
* feature(rpc): add an rpc component
* feat(rpc): add a stub for getblockchaininfo
This is the first RPC used by lightwalletd, so we need it for testing.
* fix(rpc): remove non-standard "jsonrpc: 1.0" from lightwalletd
* fix(rpc): re-enable default RPC security checks
* deps(rpc): remove not needed dependency
* fix(rpc): check if RPC task has stopped
* fix(rpc): reduce config by using Option
* fix(rpc): use tokio executor
* security(rpc): turn off rpc by default
* docs(rpc): update a TODO comment
Co-authored-by: teor <teor@riseup.net>
* fix(rpc): blocking tasks
Co-authored-by: teor <teor@riseup.net>
* rename(rpc): rpc.rs to methods.rs
* refactor(rpc): move the server to the zebra-rpc crate
* fix(rpc): clippy derive Default for RPC Config
* fix(dependencies): remove unused dependency features in zebra-rpc
We expect to use all the listed tokio features
to implement and test RPC methods.
* doc(rpc): fix testnet port, add security note
* fix(rpc): change Rust function names and update method doc TODOs
* fix(rpc): add "TODO" to fake RPC responses
* doc(rpc): update module docs
* fix(rpc): simplify server struct derives
* fix(rpc): simplify server code
* doc(rpc): explain how request fixes securely handle user-supplied data
* refactor(rpc): move the compatibility fix to a separate module
* fix(rpc): move the open log inside the spawn, and instrument it
* doc(rpc): fix toml format and provide a config example
Co-authored-by: teor <teor@riseup.net>
* fix(test): make full sync tests more efficient
Increasing the lookahead limit should increase sync speed,
at the cost of increasing memory usage.
* doc(test): remove a redundant TODO
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* feat(log): log current height when logging sync progress
* fix(test): log the specific error when full sync tests fail
* doc(start): remove an obsolete TODO
We can't decrease this log level, because the tests rely on it.
* fix(test): wait until mempool activates in full sync tests
Changes the log message and log test so that the test only finishes
when the mempool has activated.
There is still a race condition here, between the log timer and
mempool activation. But it should be very rare, because the mempool
is activated immediately when `is_close_to_tip()` becomes true.
* fix(test): warn when Zebra stalls below the maximum checkpoint height
This also improves the full sync tests,
because the warning is checked before logging a successful sync.
* feat(log): warn when sync stalls downloading the genesis block
* fix(test): warn when the state hasn't committed a block for a long time
This also improves the full sync tests,
because the warning is checked before logging a successful sync.
* doc(test): update some sync acceptance test comments
* fix(log): use Display formatting to log chrono::Duration
Debug formatting is complicated and hard to read.
* fix(log): stop saying that we've activated the mempool without checking it
We're not checking if the mempool is active, so we can't say that.
* fix(log): minor tidying and TODOs
* fix(doc): fix a typo in the tests
* fix(log): explain the post-checkpoint blocks in progress warning calculations
* fix(doc): explain what could happen if we don't wait for extra blocks
* fix(log): add a percent symbol to a percent log
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* Allow forcing colored output in `zebrad`
Add a configuration item that allows forcing Zebra to output in color
mode even if the output device is not a terminal.
* Allow enabling colored output from Zebra in tests
Force Zebrad instances to use colored output if the
`ZEBRA_FORCE_USE_COLOR` environment variable is set.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* Create a `MempoolBehavior` helper type
Prepare to replace the `Option<Height>` with a custom type that is more
flexible and provides more meaning when used.
* Use `MempoolBehavior` instead of `Option<Height>`
Clarify what the argument is, and prepare the code so that a new variant
can be added to the `MempoolBehavior` later.
* Allow the mempool to be automatically activated
Add a new variant to `MempoolBehavior` that indicates that the mempool
should become active during the test without needing a forced activation
height.
* Test full synchronization
Run `zebrad` and wait until full synchronization completes. Check if the
mempool is automatically activated when it reaches the tip.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Previously, the test would wait 10 seconds for the process to launch.
Now it waits until the process has used the conflicting resource.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* fix(network): split synthetic NotFoundRegistry from message NotFoundResponse
* docs(network): Improve `notfound` message documentation
* refactor(network): Rename MustUseOneshotSender to MustUseClientResponseSender
```
fastmod MustUseOneshotSender MustUseClientResponseSender zebra*
```
* docs(network): fix a comment typo
* refactor(network): remove generics from MustUseClientResponseSender
* refactor(network): add an inventory collector to Client, but don't use it yet
* feat(network): register missing peer responses as missing inventory
We register this missing inventory based on peer responses,
or connection errors or timeouts.
Inbound message inventory tracking requires peers to send `notfound` messages.
But `zcashd` skips `notfound` for blocks, so we can't rely on peer messages.
This missing inventory tracking works regardless of peer `notfound` messages.
* refactor(network): rename ResponseStatus to InventoryResponse
```sh
fastmod ResponseStatus InventoryResponse zebra*
```
* refactor(network): rename InventoryStatus::inner() to to_inner()
* fix(network): remove a redundant runtime.enter() in a test
* doc(network): the exact time used to filter outbound peers doesn't matter
* fix(network): handle block requests slightly more efficiently
* doc(network): fix a typo
* fmt(network): `cargo fmt` after rename ResponseStatus to InventoryResponse
* doc(test): clarify some test comments
* test(network): test synthetic notfound from connection errors and peer inventory routing
* test(network): improve inbound test diagnostics
* feat(network): add a proptest-impl feature to zebra-network
* feat(network): add a test-only connect_isolated_with_inbound function
* test(network): allow a response on the isolated peer test connection
* test(network): fix failures in test synthetic notfound
* test(network): Simplify SharedPeerError test assertions
* test(network): test synthetic notfound from partially successful requests
* test(network): MissingInventoryCollector ignores local NotFoundRegistry errors
* fix(network): decrease the inventory rotation interval
This stops us waiting 3-4 sync resets (4 minutes) before we retry a missing block.
Now we wait 1-2 sync resets (2 minutes), which is still a reasonable rate limit.
This should speed up syncing near the tip, and on testnet.
* fmt(network): cargo fmt --all
* cleanup(network): remove unnecessary allow(dead_code)
* cleanup(network): stop importing the whole sync module into tests
* doc(network): clarify syncer inventory retry constraint
* doc(network): add a TODO for a fix to ensure API behaviour remains consistent
* doc(network): fix a function doc typo
* doc(network): clarify how we handle peers that don't send `notfound`
* docs(network): clarify a test comment
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* refactor(network): rename Advertised to Available
```sh
fastmod Advertised Available zebra*
fastmod advertised available zebra*
```
* refactor(network): allow different available and missing types inside an InventoryStatus
And rename it to ResponseStatus.
Split the methods between ResponseStatus and an InventoryStatus alias.
* refactor(network): add a block_hash convenience method to InventoryHash
* test(network): improve failure logs for connection tests
* fix(inbound): move address sanitization into the response future
* feat(network): send notfound when Zebra doesn't have a block or transaction
* doc(network): move module docs to the top of each module
This makes them more likely to get updated when the module changes.
* fix(network): stop sending unsupported missing inventory types to the registry
* test(network): inbound messages are forwarded to the registry
* test(inbound): test Peers requests to the inbound service, directly and via TCP
* test(network): notfound block responses are sent by the inbound service
* test(network): notfound tx responses are sent by the inbound service
* test(network): increase sync test mock service timeout
The code that these tests use hasn't actually changed much,
and they are only failing on some platforms (coverage, macOS).
So it seems like the extra concurrent inbound tests have pushed them
past their time limit.
(Perhaps due to TCP system calls, or extra serialization work.)
* doc(network): fix typo
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* test(network): remove unnecessary multi-threaded runtime from tests
This prevents `MockService<zebra_state>` timeouts
in the `sync_block_too_high_extend_tips` test,
at the cost of reducing coverage of different execution orders.
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Log chain progress while Zebra is syncing
This helps test if the chain tip estimate is accurate,
and helps diagnose problems during full sync tests.
* Update to the latest chain tip estimate API
* Remove redundant documentation
The documentation was exactly the same as the documentation from the
trait.
* Calculate a mock time block delta for tests
Simulate a block being added to the chain with a random block time based
on the previous block time and the target spacing time.
* Add a `time` field to `ChainTipBlock`
Store the block time so that it's ready for a future chain that allows
obtaining the chain tip's block time.
* Add `ChainTip::best_tip_block_time` method
Allow obtaining the bes chain tip's block time.
* Add method to obtain both height and block time
Prevent any data races by returning both values so that they refer to
the same chain tip.
* Add `NetworkUpgrade::all_target_spacings` method
Returns all the target spacings defined for a network.
* Create a `NetworkChainTipEstimator` helper type
Isolate the code to calculate the height estimation in a new type, so
that it's easier to understand and doesn't decrease the readability of
the `chain_tip.rs` file.
* Add `ChainTip::estimate_network_chain_tip_height`
This is more of an extension method than a trait method. It uses the
`NetworkChainTipHeightEstimator` to actually perform the estimation, but
obtains the initial information from the current best chain tip.
* Fix typo in documentation
There was an extra closing bracket in the summary line.
* Refactor `MockChainTipSender` into a separate type
Prepare to allow mocking the block time of the best tip as well as the
block height.
* Allow sending mock best tip block times
Add a separate `watch` channel to send the best tip block times from a
`MockChainTipSender` to a `MockChainTip`.
The `best_tip_height_and_block_time` implementation will only return a
value if there's a height and a block time value for the best tip.
* Fix off-by-one height estimation error
Use Euclidean division to force the division result to round down
instead of rounding towards zero. This fixes an off-by-one error when
estimating a height that is lower than the current height, because the
fractionary result was being discarded, and it should have forced the
height to go one block back.
* Fix panics on local times very far in the past
Detect situations that might cause the block height estimate to
underflow, and return the genesis height instead.
* Fix another off-by-one height estimation error
The implementation of `chrono::Duration::num_seconds` adds one to the
number of seconds if it's negative. This breaks the division
calculation, so it has to be compensated for.
* Test network chain tip height estimation
Generate pairs of block heights and check that it's possible to estimate
the larger height from the smaller height and a displaced time
difference.
* Use string literal to format unreachable message
* Try removing the redundant format
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* fix(state): set state concurrency based on other services' concurrency
* fix(sync): increase the sync downloader lookahead limit
It seems like the recent tokio upgrade made this code even more efficient,
so on testnet we can have around 6000 blocks in flight.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* feat(log): log the state tip height as part of sync progress logs
* fix(log): downgrade some verbose state logs to debug
* feat(log): log successful gossiped block verification at info level
These logs help us diagnose slow progress near the tip.
There won't be very many of these logs,
because they only happen near the tip.
* fix(log): spawn top-level tasks within the global Zebra tracing span
* fix(log): spawn blocking top-level tasks within the global Zebra tracing span
Co-authored-by: teor <teor@riseup.net>
* Updating zebra-test to v1.0.0-beta.4
* Updating tower-fallback to v0.2.16
* Updating tower-batch to v0.2.20
* Updating zebra-chain to v1.0.0-beta.4
* Updating zebra-script to v1.0.0-beta.4
* Updating zebra-network to v1.0.0-beta.4
* Updating zebra-state to v1.0.0-beta.4
* Updating zebra-consensus to v1.0.0-beta.4
* Updating zebra-utils to v1.0.0-beta.4
* Updating zebrad to v1.0.0-beta.4
* Update Cargo.lock with updated crate versions
* Add section for 1.0.0-beta.4 in the CHANGELOG
* Update CHANGELOG.md
Co-authored-by: teor <teor@riseup.net>
* Update README.md
Co-authored-by: teor <teor@riseup.net>
* Update book/src/user/install.md
Co-authored-by: teor <teor@riseup.net>
* Update release description to mention Rust 2021
Update the changelog to mention that all crates now use Rust 2021
Edition.
* Elaborate on Section 3.6 documentation entry
Add the title of the section and mention that it's a section of the
Zcash protocol specification.
* Make changelog entries consistent
Use the same format for entries related to consensus rule documentation.
Co-authored-by: Marek <mail@marek.onl>
* Update README about validated consensus rules
Zebra now validates all documented consensus rules.
* Add changelog entry for newly merged PR
Describe the security fix in the changelog.
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Marek <mail@marek.onl>
* cancel background database tasks in `FinalizedState` destructor
* use `shutdown_timeout()`
* Log info-level messages while waiting for background tasks to shut down
* Cancel background tasks during debug_stop_at_height shutdown
This commit moves the database shutdown code into a common function.
* Create a constant for the tokio timeout
* Add a test script for Zebra shutdown errors
* Increase the shutdown timeout to 20 seconds for slower machines
* add title to building zebra
* use imported duration
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* Move standard lints into .cargo/config.toml
* Ignore "wrong self convention" in a futures-based trait
This lint might only trigger on beta or nightly at the moment.
* Warn if future incompatibile code is added to Zebra
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
* Use `testdir()` instead of `TempDir::new()`
Reduce repeated code and make it easier to change from using `tempdir`
to use `tempfile` instead.
* Replace `tempdir` with `tempfile` in `zebrad`
Use `tempfile`'s `TempDir` instead.
* Use `tempdir()` instead of `TempDir::new()`
Reduce repeated code and make it easier to upgrade to `tempfile`.
* Use `tempfile` instead of `tempdir`
Replace obsoleted `tempdir` dependency with `tempfile`.
* Use `tempfile` instead of `tempdir`
Replace obsoleted `tempdir` dependency with `tempfile`.
* Update `Cargo.lock`
Update it now that `tempdir` has been replaced with `tempfile`.
* Remove `tempdir` from `deny.toml` exceptions
Ban duplicate versions of the `tempdir` dependency.
* Remove `inferno` from `deny.toml` exceptions
It apparently isn't needed anymore.
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
* Refactor so that RetryLimit::Future is std::marker::Sync
* Make the syncer future std::marker::Send by spawning tips futures
* Download synced blocks in chain order, not HashSet order
* Improve MockService failure messages
* Add closure-based responses to the MockService API
* Move MockChainTip to zebra-chain
* Add a MockChainTipSender type alias
* Support MockChainTip in ChainSync and its downloader
* Add syncer tests for obtain tips, extend tips, and wrong block hashes
* Add block too high tests for obtain tips and extend tips
* Add syncer tests for duplicate FindBlocks response hashes
* Allow longer request delays for mocked services in syncer tests
* Add a copy-state command, which copies blocks between two state services
* Check blocks were written correctly
* Add extra logging to debug shutdown
* Add a block height limit argument
* Let the target state start from any height
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
* Move `MockedClientHandle` to `peer` module
It's more closely related to a `Client` than the `PeerSet`, and this
prepares it to be used by other tests.
* Rename `MockedClientHandle` to `ClientTestHarness`
Reduce confusion, and clarify that the client is not mocked.
Co-authored-by: teor <teor@riseup.net>
* Add clarification to `mock_peers` documentation
Explicitly say how the generated data is returned.
* Rename method to `wants_connection_heartbeats`
The `Client` service only represents one direction of a connection, so
`is_connected` is not the exact term.
Co-authored-by: teor <teor@riseup.net>
* Mock `Client` instead of `LoadTrackedClient`
Move where the conversion from mocked `Client` to mocked
`LoadTrackedClient` in order to make the test helper more easily used by
other tests.
* Use `ClientTestHarness` in `initialize` tests
Replace the boilerplate code to create a fake `Client` instance with
usages of the `ClientTestHarness` constructor.
* Allow receiving requests from `Client` instance
Create a helper type to wrap the result, to make it easier to assert on
specific events after trying to receive a request.
* Allow inspecting the current error in the slot
Share the `ErrorSlot` between the `Client` and the handle, so that the
handle can be used to inspect the contents of the `ErrorSlot`.
* Allow placing an error into the `ErrorSlot`
Assuming it is initially empty. If it already has an error, the code
will panic.
* Allow gracefully closing the request receiver
Close the endpoint with the appropriate call to the `close()` method.
* Allow dropping the request receiver endpoint
Forcefully closes the endpoint.
* Rename field to `client_request_receiver`
Also rename the related methods to include
`outbound_client_request_receiver` to make it more precise.
Co-authored-by: teor <teor@riseup.net>
* Allow dropping the heartbeat shutdown receiver
Allows the `Client` to detect that the channel has been closed.
* Rename fn. to `drop_heartbeat_shutdown_receiver`
Make it clear that it affects the heartbeat task.
Co-authored-by: teor <teor@riseup.net>
* Move `NowOrLater` into a new `now-or-later` crate
Make it easily accessible to other crates.
* Add `IsReady` extension trait for `Service`
Simplifies checking if a service is immediately ready to be called.
* Add extension method to check for readiness error
Checks if the `Service` isn't immediately ready because a call to
`ready` immediately returns an error.
* Rename method to `is_failed`
Avoid negated method names.
Co-authored-by: teor <teor@riseup.net>
* Add a `IsReady::is_pending` extension method
Checks if a `Service` is not ready to be called.
* Use `ClientTestHarness` in `Client` test vectors
Reduce repeated code and try to improve readability.
* Create a new `ClientTestHarnessBuilder` type
A builder to create test `Client` instances using mock data which can be
tracked and manipulated through a `ClientTestHarness`.
* Allow configuring the `Client`'s mocked version
Add a `with_version` builder method.
* Use `ClientTestHarnessBuilder` in `PeerVersions`
Use the builder to set the peer version, so that the `version` parameter
can be removed from the constructor later.
* Use a default mock version where possible
Reduce noise when setting up the harness for tests that don't really
care about the remote peer version.
* Remove `Version` parameter from the `build` method
The `with_version` builder method should be used instead.
* Fix some typos and outdated info in the release checklist
* Add extra client tests for zero and multiple readiness checks (#3273)
And document existing tests.
* Replace `NowOrLater` with `futures::poll!` (#3272)
* Replace NowOrLater with the futures::poll! macro in zebrad
* Replace NowOrLater with the futures::poll! macro in zebra-test
* Remove the now-or-later crate
* remove unused imports
* rustfmt
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Tweak crawler timings so peers are more likely to be available
* Tweak min peer connection interval so we try all peers
* Let other tasks run between fanouts, so we're more likely to choose different peers
* Let other tasks run between retries, so we're more likely to choose different peers
* Let other tasks run after peer crawler DemandDrop
This makes it more likely that peers will become ready.
* Document the chain verifier
* Drop gossiped blocks that are too far ahead of the tip
* Add extra gossiped block metrics
* Allow extra gossiped blocks, now we have a stricter limit
* Fix a comment
* Check the exact number of blocks in a downloaded block response
* Drop synced blocks that are too far ahead of the tip
* Add extra synced block metrics
* Test dropping gossiped blocks that are too far ahead of the tip
* Allow an extra checkpoint's worth of blocks in the verifier queues
* Actually let's try two extra checkpoints
* Scale extra height limit with lookahead limit
* Also drop blocks that are behind the finalized tip
* Downgrade a noisy log
* Use a debug log for already verified gossiped blocks
* Use debug logs for already verified synced blocks
* Use a named CancelHeartbeatTask unit struct for the channel type
* Prefer cancel handles in selects, if both are ready
* Fix message metrics to just show the command name
* Add metrics for internal requests and responses
* Add internal requests and responses to the messages dashboard
* Add a canceled metric, and peer addresses to request and response metrics
* Add a canceled messages graph
* Add connection state metrics for currently open connections
* Fix the connection state graph with new metrics
* Always send an error before dropping pending responses
* Move error detail logging into `fail_with`
* Delete an unused timer future
* Make error strings in metrics less verbose
* Downgrade some error logs to info
* Remove a redundant expect
* Avoid unnecessary allocations for connection state metrics
* Fix missed updates to mempool and block gossip metrics
* Stop useless crawler attempts when there are no peers and no crawl responses
* Disable GitHub bug report URLs when the disk is full
* Add help text for the `zebrad start` tracing filter option
* Use a single-thread shared Tokio runtime
This allows it to pause the time and more closely resembles the
environment that's set by default for asynchronous tests.
* Add a `zebra_test::init_async` helper function
Calls `zebra_test::init` but also constructs a single-thread Tokio
runtime and returns it. This makes it simpler to initialize asynchronous
tests that can't use the `#[tokio::test]` attribute.
* Replace usages of `Runtime::new` in tests
Use the new `zebra_test::init_async()` helper function instead.
* Replace `runtime::Builder::new_current_thread()`
Use the new `zebra_test::init_async()` helper function instead.
* Replace `runtime::Builder::new_multi_thread()`
Use the new `zebra_test::init_async()` helper function instead. The test
with the change doesn't necessarily have to use a multi-thread runtime.
* Refactor the address response limit
* Limit the number of peers in the address book
* Allow changing the address book limit in tests
* Add tests for the address book length limit
* rustfmt
* Stop checking the entire AddressBook for each connection attempt
* Stop redundant peer time checks within the address book
* Stop calling `Instant::now` 3 times for each address book update
* Only get the time once each time an address book method is called
* Update outdated comment
* Use an OrderedMap to efficiently store address book peers
* Add address book order tests
Zebra's latest beta continues implementing zero-knowledge proof and note commitment tree validation. In this release, we have finished implementing transaction header, transaction amount, and Zebra-specific NU5 validation. (NU5 mainnet validation is waiting on an `orchard` crate update, and some consensus parameter updates.)
We also fix a number of security issues that could pose a local denial of service risk, or make it easier for an attacker to make a node follow a false chain.
As of this release, Zebra will automatically download and cache the Sprout and Sapling Groth16 circuit parameters. The cache uses around 1 GB of disk space. These cached parameters are shared across all Zebra and `zcashd` instances run by the same user.
See CHANGELOG.md for the full list of changes in this release.
* Tweak a log message
* Only retry failed DNS once, then use the other DNS responses
* Limit broadcasts to half the peers
* Use a longer minimum interval for GetAddr requests
* Reduce the syncer and mempool crawler fanouts
* Stop resetting the mempool twice when it starts up
This spawns two crawlers, which send two fanouts,
so it can use up a lot of peers.
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Download and load Sprout parameters using zcash_proofs
Also update some librustzcash dependencies, to avoid duplicate dependencies.
* Update upstream orchard to avoid a compilation error
* Skip librustzcash batch refactor for now, to avoid compilation errors
* Change the cache ID, so we actually cache Sprout
* Move existing file checks into zcash_proofs
* Add a 1 hour timeout to parameter file downloads
* Give other tasks priority, before spawning the download task
* Update to the latest version of our modified librustzcash fork
* Change the cache key for Sprout
* Add 40 minutes to CI timeouts for occasional sprout downloads
* Update to zcash_proofs with split downloads
* Check file sizes to help debug parameter load failures in zcash_proofs
* Start the second download once the first has finished in zcash_proofs
* Document the parameter download task
* Stop hashing existing files twice
* Start network before verifiers
This makes the Groth16 download task start as late as possible.
* Explain why the Groth16 download must happen first
* Speed up Zebra shutdown: skip waiting for the tokio runtime
* Replace Zcash parameters crates with pre-downloaded local parameter files
* Download Zcash parameters using the `zcashd` script in CI and Docker
* Add a zcash_proofs dependency to zebra-consensus
* Download Sapling parameters using zcash_proofs, rather than fetch-params.sh
* Add a new `zebrad download` subcommand
This command isn't required for nomrmal usage.
But it's useful when testing, or launching multiple Zebra instances.
* Use `zebrad download` in CI to pre-download parameters
* Log a helpful hint if downloading fails
* Allow some duplicate dependencies currently hidden by orchard
* Spawn a separate task to download Groth16 parameters
* Run the parameter download with code coverage
This avoids re-compining Zebra with and without coverage.
* Update Cargo.lock after rebase
* Try to pass `download` as an argument to `zebrad` in coverage CI
* Fix copy and paste comment typos
* Add path and download examples, like zcash_proofs
* Download params in CI just like zcash_proofs does
* Delete a redundant build step
* Implement graceful shutdown for zebrad start
* Send coverage summary to /dev/null when getting the params path
* Use the correct parameters path and download commands in CI
* Explain pre-downloads
* Avoid calling params_folder twice
* Rename parameter types and methods for consistency
```sh
fastmod SaplingParams SaplingParameters zebra*
fastmod Groth16Params Groth16Parameters zebra*
fastmod PARAMS GROTH16_PARAMETERS zebra*
fastmod params_folder directory zebra*
```
And a manual variable name tweak.
* rustfmt
* Remove a redundant coverage step
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Implement graceful shutdown for the peer set
* Use the minimum lookahead limit in acceptance tests
* Enable a doctest that compiles with newly public modules
* Revert "Remove commented-out code"
This reverts commit 9e69777925f103ee11e5940bba95b896c828839b.
* Implement deserialization for `addrv2` messages
* Limit addr and addrv2 messages to MAX_ADDRS_IN_MESSAGE
* Clarify address version comments
* Minor cleanups and fixes
* Add preallocation tests for AddrV2
* Add serialization tests for AddrV2
* Use prop_assert in AddrV2 proptests
* Use a generic utility method for deserializing IP addresses in `addrv2`
* Document the purpose of a conversion to MetaAddr
* Fix a comment typo, and clarify that comment
* Clarify the unsupported AddrV2 network ID error and enum variant names
```sh
fastmod AddrV2UnimplementedError UnsupportedAddrV2NetworkIdError zebra-network
fastmod Unimplemented Unsupported zebra-network
```
* Fix and clarify unsupported AddrV2 comments
* Replace `panic!` with `unreachable!`
* Clarify a comment about skipping a length check in a test
* Remove a redundant test
* Basic addr (v1) and addrv2 deserialization tests
* Test deserialized IPv4 and IPv6 values in addr messages
* Remove redundant io::Cursor
* Add comments with expected values of address test vectors
* Update `tower` to version `0.4.9`
Update to latest version to add support for Tokio version 1.
* Replace usage of `ServiceExt::ready_and`
It was deprecated in favor of `ServiceExt::ready`.
* Update Tokio dependency to version `1.13.0`
This will break the build because the code isn't ready for the update,
but future commits will fix the issues.
* Replace import of `tokio::stream::StreamExt`
Use `futures::stream::StreamExt` instead, because newer versions of
Tokio don't have the `stream` feature.
* Use `IntervalStream` in `zebra-network`
In newer versions of Tokio `Interval` doesn't implement `Stream`, so the
wrapper types from `tokio-stream` have to be used instead.
* Use `IntervalStream` in `inventory_registry`
In newer versions of Tokio the `Interval` type doesn't implement
`Stream`, so `tokio_stream::wrappers::IntervalStream` has to be used
instead.
* Use `BroadcastStream` in `inventory_registry`
In newer versions of Tokio `broadcast::Receiver` doesn't implement
`Stream`, so `tokio_stream::wrappers::BroadcastStream` instead. This
also requires changing the error type that is used.
* Handle `Semaphore::acquire` error in `tower-batch`
Newer versions of Tokio can return an error if the semaphore is closed.
This shouldn't happen in `tower-batch` because the semaphore is never
closed.
* Handle `Semaphore::acquire` error in `zebrad` test
On newer versions of Tokio `Semaphore::acquire` can return an error if
the semaphore is closed. This shouldn't happen in the test because the
semaphore is never closed.
* Update some `zebra-network` dependencies
Use versions compatible with Tokio version 1.
* Upgrade Hyper to version 0.14
Use a version that supports Tokio version 1.
* Update `metrics` dependency to version 0.17
And also update the `metrics-exporter-prometheus` to version 0.6.1.
These updates are to make sure Tokio 1 is supported.
* Use `f64` as the histogram data type
`u64` isn't supported as the histogram data type in newer versions of
`metrics`.
* Update the initialization of the metrics component
Make it compatible with the new version of `metrics`.
* Simplify build version counter
Remove all constants and use the new `metrics::incement_counter!` macro.
* Change metrics output line to match on
The snapshot string isn't included in the newer version of
`metrics-exporter-prometheus`.
* Update `sentry` to version 0.23.0
Use a version compatible with Tokio version 1.
* Remove usage of `TracingIntegration`
This seems to not be available from `sentry-tracing` anymore, so it
needs to be replaced.
* Add sentry layer to tracing initialization
This seems like the replacement for `TracingIntegration`.
* Remove unnecessary conversion
Suggested by a Clippy lint.
* Update Cargo lock file
Apply all of the updates to dependencies.
* Ban duplicate tokio dependencies
Also ban git sources for tokio dependencies.
* Stop allowing sentry-tracing git repository in `deny.toml`
* Allow remaining duplicates after the tokio upgrade
* Use C: drive for CI build output on Windows
GitHub Actions uses a Windows image with two disk drives, and the
default D: drive is smaller than the C: drive. Zebra currently uses a
lot of space to build, so it has to use the C: drive to avoid CI build
failures because of insufficient space.
Co-authored-by: teor <teor@riseup.net>
* Fix some links and linewraps
* Flesh out module-level rustdoc for mempool component
* Links for Downloads and Storage
* Tidy rustdoc links
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Avoid spurious test failures on busy machines
* Revert "Avoid spurious test failures on busy machines"
This reverts commit 4430a48ef3d20ceddf737a2f2741638875b0148c.
* Increase request delay for mock network service
This is necessary because some tests were having timeouts when running
on macOS for the CI. This meant that it took longer than expected for
the subject under test to end up sending the network request.
Co-authored-by: teor <teor@riseup.net>
* Tidy links and notes for some mempool rustdoc
* Fix link
* Fix some links and linewraps
* minor spacing
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Limit the size and age of the ZIP-401 rejected transaction ID list
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* Fix bug in EvictionList; improve documentation
* Separate public and non-public parts of the documentation
* Fix tests
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* Fix bug in EvictionList::len()
* Make EvictionList::len() mutable; prune the list inside it
* Limit the size of EvictedList::ordered_entries
* Increase eviction_list_time_mixed time constants to try to make it pass on MacOS
* Simplify logic by assuming refreshes will never happen
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* Compiling fixes
* Remove MEMPOOL_SIZE and just rely on the ZIP-401 cost limit
Co-authored-by: teor <teor@riseup.net>
* ZIP-401 weighted random mempool eviction
* rename zcash.mempool.total_cost.bytes to zcash.mempool.cost.bytes
Co-authored-by: teor <teor@riseup.net>
* Remove duplicated lines
* Add cost() method to UnminedTx
Update serialization failure messages
* More docs quoting ZIP-401 rules
* Change mempool::Storage::new() to handle Copy-less HashMap, HashSet
* mempool: tidy cost types and evict_one()
* More consensus rule docs
* Refactor calculating mempool costs for Unmined transactions
* Add a note on asympotic performance of calculating weights of txs in mempool
* Bump test mempool / storage config to avoid weighted random cost limits
* Use mempool tx_cost_limit = u64::MAX for some tests
* Remove failing tests for now
* Allow(clippy::field-reassign-with-default) because of a move on a type that doesn't impl Copy
* Fix mistaken doctest formatting
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Increase test timeout for Windows builds
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Try simulating a chain growth
* Adjust the transaction expiry height
The mempool evicts expired transactions. When working with mocked data,
appending a new block typically clears the mempool because transactions become
expired. For this reason, the expiry height of each transactions is adjusted so
that it is greater than the new chain tip's height.
* Refactor the code so that it works with `VerifiedUnminedTx`
* Fix a typo
* Fix clippy warnings
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
There are a lot of these messages when Zebra starts up.
They might be slowing down CI and causing timeouts.
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
This matches the settings for `sync_large_checkpoints_mainnet`.
Also reduce the number of blocks synced to reduce network load.
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Increment the crates that have new commits since the last version
* Increment the crates that depend on crates that have changed
* Increment the version of `zebra-script`
* Use the `zebrad` version in the `zebra-network` user agent string
* Use the `v1.0.0-alpha.19` git tag in `README.md`
* Copy the draft changelog into `CHANGELOG.md`
* Delete bumps
* Update CHANGELOG.md
Co-authored-by: teor <teor@riseup.net>
* Add newly merged PRs
Co-authored-by: teor <teor@riseup.net>
* Create a `NowOrLater` helper type
A replacement for `FutureExt::now_or_never` that ensures that the task
is scheduled for waking up later when the inner future is ready.
* Use `NowOrLater` to fix possible delay bug
Previous usage of `now_or_never` meant that the underlying task wasn't
being scheduled to awake when the `Downloads` stream produced a new
item. Using `NowOrLater` instead fixes that issue.
* Increase the restart test timeout to 10 seconds
It shouldn't take this long.
But maybe the CI VMs are under a lot of load?
* Add extensive logging to diagnose CI state reload failures
* Ignore AlreadyInChain error in the syncer
* Split Cancelled errors; add them to should_restart_sync exceptions
* Also filter 'block is already comitted'; try to detect a wrong downcast
* Rename tx downloader & verifier metrics
* Add version to mempool metrics
* Add new metrics
* Make sure mempool gauges are zeroed when instances are dropped
* Updated mempool grafana dashboard
* Removed transaction verification dashboard; moved to mempool
* Update mempool dashboard
* Add reason to error labels in mempool dashboard
* Rename some metrics per review
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Guarantee unique IDs in mempool service responses
* Guarantee unique IDs in crawler task mempool Queue requests
Also update the tests to use unique IDs.
* Add a CheckForVerifiedTransactions mempool request
Also document the mempool request and response variants.
* Spawn a QueueChecker task to check for newly verified transactions
This task makes sure that transactions reliably propagate,
rather than relying on peer requests or responses to trigger propagation.
* Update the start command documentation
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Impl Drop, Default and take() for ActiveState
* Refactor Mempool::poll_ready to check disabled and reset first
Also remove some levels of nesting.
* Use the same code for dropping and resetting the mempool
* Document where the tasks are dropped when switching states
* Log mempool resets at info level
And add heights to mempool enable/disable/reset logs
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Guarantee unique IDs in mempool service responses
* Guarantee unique IDs in crawler task mempool Queue requests
Also update the tests to use unique IDs.
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Create a new VerifiedUnminedTx containing the miner fee
* Use VerifiedUnminedTx in mempool verification responses
And do a bunch of other cleanups.
* Use VerifiedUnminedTx in mempool download and verifier
* Use VerifiedUnminedTx in mempool storage and verified set
* Impl Display for VerifiedUnminedTx, and some convenience methods
* Use VerifiedUnminedTx in existing tests
* add some additional checks to the acceptance mempool test
* add an additional mempool test
* do proposed fixes to `sync_until`
* Ignore "can't kill an exited process" errors
Co-authored-by: teor <teor@riseup.net>
* Get the transaction fee from utxos
* Return the transaction fee from the verifier
* Avoid calculating the fee for coinbase transactions
Coinbase transactions don't have fees. In case of a coinbase transaction, the
verifier returns a zero fee.
* Update the result obtained by `Downloads`
* Update some comments
* Add a mempool debug_enable_at_height config
* Rename a field in the mempool crawler
* Propagate syncer channel errors through the crawler
We don't want to ignore these errors, because they might indicate a shutdown.
(Or a bug that we should fix.)
* Use debug_enable_at_height in the mempool crawler
* Log when the mempool is activated or deactivated
* Deny unknown fields and apply defaults for all configs
* Move Duration last, as required for TOML tables
* Add a basic mempool acceptance test
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* do not advertise rejected transactions
* do not broadcast transaction that are expired
* change dummy var name
* simplify code, performance
* clippy
* add some test coverage
* clippy
Co-authored-by: teor <teor@riseup.net>
* Split mempool config into its own module
Also:
- expand config docs
- clean up mempool imports
* Pass the mempool config to the mempool
* Create the transaction sender channel inside the mempool 1/2
This simplifies all the code that calls the mempool.
Also:
- update the mempool enabled state before returning the new mempool
- add some test module doc comments
* Refactor a setup function out of the mempool unit tests 2/2
Also:
- update the setup function to handle the latest mempool changes
* Clarify a comment
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Limit the size of rejection lists when there is a spend conflict
Previously, `insert` would return early with an error,
and skip limiting the rejection list sizes.
* Use prop_assert macros in proptests, rather than assert
* Add `HashSet`s to help spend conflict detection
Keep track of the spent transparent outpoints and the revealed
nullifiers.
Clippy complained that the `ActiveState` had variants with large size
differences, but that was expected, so I disabled that lint on that
`enum`.
* Clear the `HashSet`s when clearing the mempool
Clear them so that they remain consistent with the set of verified
transactions.
* Use `HashSet`s to check for spend conflicts
Store new outputs into its respective `HashSet`, and abort if a
duplicate output is found.
* Remove inserted outputs when aborting
Restore the `HashSet` to its previous state.
* Remove tracked outputs when removing a transaction
Keep the mempool storage in a consistent state when a transaction is
removed.
* Remove tracked outputs when evicting from mempool
Ensure eviction also keeps the tracked outputs consistent with the
verified transactions.
* Refactor to create a `VerifiedSet` helper type
Move the code to handle the output caches into the new type. Also move
the eviction code to make things a little simpler.
* Refactor to have a single `remove` method
Centralize the code that handles the removal of a transaction to avoid
mistakes.
* Move mempool size limiting back to `Storage`
Because the evicted transactions must be added to the rejected list.
* Remove leftover `dbg!` statement
Leftover from some temporary testing code.
Co-authored-by: teor <teor@riseup.net>
* Remove unnecessary `TODO`
It is more speculation than planning, so it doesn't add much value.
Co-authored-by: teor <teor@riseup.net>
* Fix typo in documentation
The verb should match the subject "transactions" which is plural.
Co-authored-by: teor <teor@riseup.net>
* Add a comment to warn about correctness
There's a subtle but important detail in the implementation that should
be made more visible to avoid mistakes in the future.
Co-authored-by: teor <teor@riseup.net>
* Remove outdated comment
Left-over from the attempt to move the eviction into the `VerifiedSet`.
* Improve comment explaining lint removal
Rewrite the comment explaining why the Clippy lint was ignored.
* Check for spend conflicts in `VerifiedSet`
Refactor to avoid API misuse.
* Test rejected transaction rollback
Using two transactions, perform the same test adding a conflict to both
of them to check if the second inserted transaction is properly
rejected. Then remove any conflicts from the second transaction and add
it again. That should work, because if it doesn't it means that when the
second transaction was rejected it left things it shouldn't in the
cache.
* Test removal of multiple transactions
When removing multiple transactions from the mempool storage, all of the
ones requested should be removed and any other transaction should be
still be there afterwards.
* Increase mempool size to 4, so that spend conflict tests work
If the mempool size is smaller than 4,
these tests don't fail on a trivial removal bug.
Because we need a minimum number of transactions in the mempool
to trigger the bug.
Also commit a proptest seed that fails on a trivial removal bug.
(This seed fails if we remove indexes in order,
because every index past the first removes the wrong transaction.)
* Summarise transaction data in proptest error output
* Summarise spend conflict field data in proptest error output
* Summarise multiple removal field data in proptest error output
And replace the very large proptest debug output with the new summary.
Co-authored-by: teor <teor@riseup.net>
* bradcast transactions to peers after they get inserted into mempool
* remove network argument from mempool init
* remove dbg left
* remove return value in mempool enable call
* rename channel sender and receiver vars
* change unwrap() to expect()
* change the channel to a hashset
* fix build
* fix tests
* rustfmt
* fix tiny space issue inside macro
Co-authored-by: teor <teor@riseup.net>
* check errors/panics in transaction gossip tests
* fix build of newly added tests
* Stop dropping the inbound service and mempool in a test
Keeping the mempool around avoids a transaction broadcast task error,
so we can test that there are no other errors in the task.
* Tweak variable names and add comments
* Avoid unexpected drops by returning a mempool guard in tests
* Use BoxError to simplify service types in tests
* Make all returned service types consistent in tests
We want to be able to change the setup without changing the tests.
Co-authored-by: teor <teor@riseup.net>
* Split mempool storage errors into tip-based and chain-based
* Expire tip rejections every time we get a new block
FailedVerification and SpendConflict rejections only apply to the current tip.
The next tip can provide missing inputs, or evict conflicting transactions.
* Enforce MAX_EVICTION_MEMORY_ENTRIES for mempool reject lists
* Implement a task that gossips verified block hashes
* Log an info message for block broadcasts
* Simplify the gossip task
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Re-use the old tip change if there is no new tip change
Also improve the comments.
* Add an assertion message
* Rename task join handles and futures in start method
* Add a dedicated BlockGossipError type
This type helps distinguish between syncer and state errors.
* Test that committed blocks are gossiped to peers
Also do a minor type cleanup on the existing test code,
replacing `Option<Vec<_>>` with `Vec<_>`.
* Formatting
* Remove excess newlines
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Clear the initial gossiped blocks during test setup
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Remove unused mempool storage errors
Preparation for ticket #2819.
Removing these errors means that we don't have to decide
which type of transaction ID match we want for them.
* Remove unused mempool errors, and deduplicate storage errors
* rustfmt
* Add a mempool transaction removal method for mined IDs
And use this method to remove expired transactions,
because all transactions with the same mined ID expire at the same height.
* Remove mined transaction IDs from the mempool
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Update versions for zebra v1.0.0-alpha.18 release
* WIP: Initial PR list
* Remove uninteresting version bumps from CHANGELOG
* Categorise and group PRs in CHANGELOG, removing uninteresting PRs
* Further refine and categorise changelog entries
* Fix tag url
* Final changes to CHANGELOG
* Add a changelog description
* Spacing
* Clarify and fix changelog PR descriptions
* Add PRs that are about to be merged
* More slight clarifications
* Spacing
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Rename type parameter to be more explicit
Replace the single letter with a proper name.
* Remove imports for `Request` and `Response`
The type names will conflict with the ones for the mempool service.
* Attach `Mempool` service to the `Crawler`
Add a field to the `Crawler` type to store a way to access the `Mempool`
service.
* Forward crawled transactions to downloader
The crawled transactions are now sent to the transaction downloader and
verifier, to be included in the mempool.
* Derive `Eq` and `PartialEq` for `mempool::Request`
Make it simpler to use the `MockService::expect_request` method.
* Test if crawled transactions are downloaded
Create some dummy crawled transactions, and let the crawler discover
them. Then check if they are forwarded to the mempool to be downloaded
and verified.
* Don't send empty transaction ID list to downloader
Ignore response from peers that don't provide any crawled transactions.
* Log errors when forwarding crawled transaction IDs
Calling the Mempool service should not fail, so if an error happens it
should be visible. However, errors when downloading individual
transactions can happen from time to time, so there's no need for them
to be very visible.
* Document existing `mempool::Crawler` test
Provide some depth as to what the test expect from the crawler's
behavior.
* Refactor to create `setup_crawler` helper function
Make it easier to reuse the common test setup code.
* Simplify code to expect requests
Now that `zebra_network::Request` implement `Eq`, the call can be
simplified into `expect_request`.
* Refactor to create `respond_with_transaction_ids`
A helper function that checks for a network crawl request and responds
with the given list of crawled transaction IDs.
* Refactor to create `crawler_iterator` helper
A function to intercept and respond to the fanned-out requests sent
during a single crawl iteration.
* Refactor to create `respond_to_queue_request`
Reduce the repeated code necessary to intercept and reply to a request
for queuing transactions to be downloaded.
* Add `respond_to_queue_request_with_error` helper
Intercepts a mempool request to queue transactions to be downloaded, and
responds with an error, simulating an internal problem in the mempool
service implementation.
* Derive `Arbitrary` for `NetworkUpgrade`
This is required for deriving `Arbitrary` for some error types.
* Derive `Arbitrary` for `TransactionError`
Allow random transaction errors to be generated for property tests.
* Derive `Arbitrary` for `MempoolError`
Allow random Mempool errors to be generated for property tests.
* Test if errors don't stop the mempool crawler
The crawler should be robust enough to continue operating even if the
mempool service fails to download transactions or even fails to handle
requests to enqueue transactions.
* Reduce the log level for download errors
They should happen regularly, so there's no need to have them with a
high visibility level.
Co-authored-by: teor <teor@riseup.net>
* Stop crawler if service stops
If `Mempool::poll_ready` returns an error, it's because the mempool
service has stopped and can't handle any requests, so the crawler should
stop as well.
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Move mempool tests into `tests::vector` sub-module
Make it consistent with other test modules and prepare for adding
property tests.
* Reorder imports
Make it consistent with the general guidelines followed on other
modules.
* Export `ChainTipBlock` and `ChainTipSender`
Allow these types to be used in other crates for testing purposes.
* Derive `Arbitrary` for `ChainTipBlock`
Make it easy to generate random `ChainTipBlock`s for usage in property
tests.
* Refactor to move test methods into `tests` module
Reduce the repeated test configuration attributes and make it easier to
see what is test specific and what is part of the general
implementation.
* Add a `Mempool::dummy_call` test helper method
Performs a dummy call just so that `poll_ready` gets called.
* Use `dummy_call` in existing tests
Replace the custom dummy requests with the helper method.
* Test if the mempool is cleared on chain reset
A chain reset should force the mempool storage to be cleared so that
transaction verification can restart using the new chain tip.
* Test if mempool is cleared on syncer restart
If the block synchronizer falls behind and then starts catching up
again, the mempool should be disabled and therefore the storage should
be cleared.
* Send mined transaction IDs to the download and verify task for cancellation
* Pass a HashSet of transaction hashes to be cancelled
* Add mempool_cancel_mined() test
* Fix starvation in test
* Fix typo in comment
* mempool - support transaction expiration
* use `LatestChainTip` instead of state call
* clippy
* remove spawn task
* remove non needed async from function
* remove return value
* add a `expiry_height_mut()` method to `Transaction` for testing purposes
* fix `remove_expired_transactions()`
* add a `mempool_transaction_expiration()` test
* tidy cleanup to `expiry_height()`
* improve docs
* fix the build
* try fix macos build
* extend tests
* add doc to function
* clippy
* fix build
* start tests at block two
* Cancel download and verify tasks when the mempool is deactivated
* Refactor enable/disable logic to use a state enum
* Add helper test functions to enable/disable the mempool
* Add documentation about errors on service calls
* Improvements from review
* Improve documentation
* Fix bug in test
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: teor <teor@riseup.net>
* Add `Transaction::spent_outpoints` getter method
Returns an iterator over the UTXO `OutPoint`s spent by the transaction.
* Add `mempool::Error::Conflict` variant
An error representing that a transaction was rejected because it
conflicts with another transaction that's already in the mempool.
* Reject conflicting mempool transactions
Reject including a transaction in the mempool if it spends outputs
already spent by, or reveals nullifiers already revealed by another
transaction in the mempool.
* Fix typo in documentation
Remove the `r` that was incorrectly added.
Co-authored-by: teor <teor@riseup.net>
* Specify that the conflict is a spend conflict
Make the situation clearer, because there are other types of conflict.
Co-authored-by: teor <teor@riseup.net>
* Clarify that the outpoints are from inputs
Because otherwise it could lead to confusion because it could also mean
the outputs of the transaction represented as `OutPoint` references.
Co-authored-by: teor <teor@riseup.net>
* Create `storage::tests::vectors` module
Refactor to follow the convention used for other tests.
* Add an `AtLeastOne::first_mut` method
A getter to allow changing the first element.
* Add an `AtLeastOne::push` method
Allow appending elements to the collection.
* Derive `Arbitrary` for `FieldNotPresent`
This is just to make the code that generates arbitrary anchors a bit
simpler.
* Test if conflicting transactions are rejected
Generate two transactions (either V4 or V5) and insert a conflicting
spend, which can be either a transparent UTXO, or a nullifier for one of
the shielded pools. Check that any attempt to insert both transactions
causes one to be accepted and the other to be rejected.
* Delete a TODO comment that we decided not to do
Co-authored-by: teor <teor@riseup.net>
* Add tests for mempool Request::Queue
* Update test to work after refactoring
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
* Send `Response::Nil` instead of sending empty `Message`s
This matches `zcashd`'s behaviour more closely.
In most cases, the network layer filters these out already.
But this change makes the the inbound service code clearer.
* revert changes made to `AdvertiseTransactionIds` and `PushTransaction`
* remove newline
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* Update the expiry TODO
* Clear the mempool at a chain tip reset
* Clear the mempool by using a sync method (#2777)
* Clear the mempool by using a sync method
* Update docs
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* Refactor last_tip_change()
* Apply suggestions from code review
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Fix brackets
* Use best_tip_block instead of manual borrowing
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Remove return of redundant vector length
An attempt to improve readability a bit by not returning a tuple with a
value that can be obtained from a single return type.
* Refactor `unmined_transactions_in_blocks`
Use a more functional style to try to make it a bit clearer.
* Use ranges in `unmined_transactions_in_blocks`
Allow a finer control over the block range to extract the transactions
from.
* Refactor mempool test code to improve clarity
It was previously not clear that only the first genesis transaction was
being used. The remaining transactions in the genesis block were
discarded and then readded later.
* Replace `oneshot` with `call`
Remove a redundant internal `ready_and` call.
* Return an `impl Iterator` instead of a `Vec<_>`
Remove unnecessary deserializations and heap allocations.
* Refactor `mempool_storage_basic_for_network` test
Make the separation between the transactions expected to be in the
mempool and those expected to be rejected clearer.
* Replace `Iterator` with `DoubleEndedIterator`
Allow getting the last transaction more easily.
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Use `MockService` in inbound test
Refactor the `mempool_requsets_for_transactions` test so that it uses a
`MockService` instead of the `mock_peer_set` function.
* Use `MockService` in the basic mempool test
Refactor the `mempool_service_basic` test so that it uses a
`MockService` instead of the `mock_peer_set` helper function.
* Remove the `mock_peer_set` helper function
It is not used anymore, since the usages were replaced with
`MockService`s.
* add tests for mempool inbound requests
* Use MockService for transaction verifier
* Refactor creation of mock `peer_set`
Use the same style as the mock transaction verifier.
* Derive `Eq` for `zebra_network::Request`
Make it easy to use the `MockService::expect_request` method.
* Return mocked peer set service from `setup`
Allow it to be used to respond to requests.
* Add bindings for the transaction used for testing
Allow them to be moved into futures later.
* Respond to transaction download request
Make sure that the test transaction appears to the mempool as if it had
been downloaded by the peer set service.
* Assert that no unexpected requests were received
Check that the mempool doesn't send unexpected requests to the peer set
service.
* add tests for mempool inbound requests
* Use MockService for transaction verifier
* add missing `expect_no_requests` to `mempool_advertise_transaction_ids` test
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Use `MockService` in inbound test
Refactor the `mempool_requsets_for_transactions` test so that it uses a
`MockService` instead of the `mock_peer_set` function.
* Use `MockService` in the basic mempool test
Refactor the `mempool_service_basic` test so that it uses a
`MockService` instead of the `mock_peer_set` helper function.
* Remove the `mock_peer_set` helper function
It is not used anymore, since the usages were replaced with
`MockService`s.
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Allow deliberate instances of the new nightly clippy::derivable_impls lint
We want our config defaults to be explicit.
Not so sure about the application defaults, but they also contain a config.
* Also allow unknown lint names
Stable doesn't know about this lint, but nightly does.
* Implement initial service mocking helpers
Adds a [`MockService`] type, which can be configured and built for usage
in unit tests or proptests. The mocked service can then be used to
intercept requests and respond indivdiually to them.
* Use `MockService in the `mempool::Crawler` test
Refactor it to remove the helper mock function, and use the new
`MockService` helper type.
* Use `MockService` in `CandidateSet` test vectors
Refactor to remove the manual mocking of the peer set service.
* Panic if a response is not sent by `MockService`
Change the current semantics to require all `MockService` usages to
respond to every intercepted request.
A `must_use` attribute was added to the `ResponseSender` so that the
compiler can warn when this doesn't happen.
* Allow generic error types in `MockService`
Replace the hard-coded `BoxError` as the `Service`'s error type with a
generic type parameter. This allows mocking services in locations that
require specific error types.
* Add a `ResponseSender::request` getter
Allow inspecting the request again before responding, and using
information from the request in the response.
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Pass sync_status to mempool
* Update zebrad/src/components/mempool.rs
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
* Remove enabled flag for now; will be handled in #2723
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
* Check if tx already exists in mempool or state before downloading
* Reorder checks
* Add rejected test; refactor into separate function
* Wrap mempool in buffered service
* Rename RejectedTransactionsById -> RejectedTransactionsIds
* Add RejectedTransactionIds response; fix request name
* Organize imports
* add a test for Storage::rejected_transactions
* add test for mempool `Request::RejectedTransactionIds`
* change buffer size to 1 in the test
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
Using `&mut self` as the receiver in the method signatures allows Rust
to infer that the type is properly `Sync`, and therefore `Send`. This
allows removing the `Mutex` work-around.
* Decide if Zebra is at the chain tip
* Avoid division by zero
* Try increasing EVENT_TIMEOUT
* Increase MAX_TEST_EXECUTION
* Implement basic tests
* Resolve Clippy's erorrs
* change doc comments to normal
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
* reply to `Request::MempoolTransactionIds`
* remove boilerplate
* get storage from mempool with a method
* change panic message
* try fix for mac
* use normal init instead of init_tests for state service
* newline
* rustfmt
* fix test build
* Rename ChainTipReceiver to CurrentChainTip
`fastmod ChainTipReceiver CurrentChainTip zebra*`
* Update chain tip documentation and variable names
* Basic chain tip change implementation, without resets
Also includes the following name changes:
```
fastmod CurrentChainTip LatestChainTip zebra*
fastmod chain_tip_receiver latest_chain_tip zebra*
```
* Clarify the difference between `LatestChainTip` and `ChainTipChange`
* Store a `SyncStatus` handle in the `Crawler`
The helper type will make it easier to determine if the crawler is
enabled or not.
* Pause crawler if mempool is disabled
Implement waiting until the mempool becomes enabled, so that the crawler
does not run while the mempool is disabled.
If the `MempoolStatus` helper is unable to determine if the mempool is
enabled, stop the crawler task entirely.
* Update test to consider when crawler is paused
Change the mempool crawler test so that it's a proptest that tests
different chain sync. lengths. This leads to different scenarios with
the crawler pausing and resuming.
Co-authored-by: teor <teor@riseup.net>
* Create a `SyncStatus` helper type
Keeps track if the synchronizer is close to the chain tip or not.
* Refactor `ChainSync` ctor. to return `SyncStatus`
Change the constructor API so that it returns a higher level construct.
* Test if `SyncStatus` waits for the chain tip
Test if waiting for the chain tip to be reached correctly finishes when
the chain tip is reached. This is done by sending recent sync lengths to
the `SyncStatus` instance, and checking that every time a separate
`SyncStatus` instance determines it has reached the tip the original
instance wakes up.
* Add a temporary attribute to allow dead code
The code added isn't used yet, so we'll add a temporary waiver until
another PR is merged to use them.
This avoids peer set contention when most peers are busy.
Also exit the task if the peer service returns a readiness error,
because that means it's permanently unusable.
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* First pass at a Mempool Service, incl. a storage layer underneath
* Fixed up Mempool service and storage
* allow dead code where needed
* clippy
* typo
* only drain if the mempool is full
* add a basic storage test
* remove space
* fix test for when MEMPOOL_SIZE change
* group some imports
* add a basic mempool service test
* add clippy suggestions
* remove not needed allow dead code
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
Co-authored-by: teor <teor@riseup.net>
* Rename BestTipHeight so it can be generalised to ChainTipSender
`fastmod BestTipHeight ChainTipSender zebra*`
For senders:
`fastmod best_tip_height chain_tip_sender zebra*`
For receivers:
`fastmod best_tip_height chain_tip_receiver zebra*`
* Rename best_tip_height module to chain_tip
* Wrap the chain tip watch channel in a ChainTipReceiver type
* Create a ChainTip trait to avoid tricky crate dependencies
And add convenience impls for optional and empty chain tips.
* Use the ChainTip trait in zebra-network
* Replace `Option<ChainTip>` with `NoChainTip`
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Return a transaction verifier from `zebra_consensus::init`
This verifier is temporarily created separately from the block verifier's
transaction verifier.
* Return the same transaction verifier used by the block verifier
* Clarify that the mempool verifier is the transaction verifier
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
Co-authored-by: Deirdre Connolly <deirdre@zfnd.org>
Co-authored-by: Conrado Gouvea <conrado@zfnd.org>
* Create initial `mempool::Crawler` type
The mempool crawler is responsible for periodically asking peers for
transactions to insert into the local mempool. This initial
implementation will periodically ask for transactions, but won't do
anything with them yet.
Also, the crawler is currently configured to be always enabled, but this
should be fixed to avoid crawling while Zebra is still syncing the
chain.
* Add a timeout to peer responses
Prevent the crawler from getting stuck if there's communication with a
peer that takes too long to respond.
* Run the mempool crawler in Zebra
Spawn a task for the crawler when Zebra starts.
* Test if the crawler is sending requests
Create a mock for the `PeerSet` service to intercept requests and verify
that the transaction requests are sent periodically.
* Use `full` Tokio features when testing
Make it simpler to select the features for test builds.
Co-authored-by: teor <teor@riseup.net>
* Link to the issue for crawler activation
Make it easy to navigate from the `TODO` comment to the current project
planning.
Co-authored-by: teor <teor@riseup.net>
* Link to the issue for downloading transactions
Make it easy to navigate from the `TODO` comment to the current project
planning.
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: teor <teor@riseup.net>
* Minimal recent sync lengths implementation
Also includes metrics and logging, to make diagnosing bugs easier.
* Add logging to check what happens when Zebra reaches the chain tip
* Add tests for recent sync lengths
- initially empty
- pruned to correct length
- newest entries go first
* Drop a redundant `/` from a Cargo.lock URL
This seems to be a nightly or beta Rust change,
but hopefully stable just accepts it.
* Use metrics histograms to avoid overwriting values
* Add detailed syncer monitoring dashboard
* Increase the recent sync length to 4
This length makes it easier to distinguish between temporary and
sustained errors/syncs.
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Rename internal network requests for wide transaction IDs
fastmod TransactionsByHash TransactionsById zebra*
fastmod AdvertiseTransactions AdvertiseTransactionIds zebra*
fastmod MempoolTransactions MempoolTransactionIds zebra*
fastmod TransactionHashes TransactionIds zebra*
* Update network transaction request/response comments
* Rename a transaction hash method for wide transaction IDs
fastmod transaction_hashes transaction_ids zebra-network
* Add UnminedTxId methods and conversions for InventoryHash
* Map WtxIds to unmined transaction network messages
Also, use UnminedTxId and UnminedTx in:
* Zebra's internal request and response format, and
* external Zcash network protocol messages.
* Enable WtxId mempool inventory tracking for peers
* Further clarify transaction IDs
* Use Witnessed rather than Wide for transaction IDs
And rename narrow to legacy when it only applies to v1-v4 transactions.
Otherwise, rename it to mined ID.
* Rename a missed binding
* Remove an incorrectly named binding
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Simplify state service initialization in test
Use the test helper function to remove redundant code.
* Create `BestTipHeight` helper type
This type abstracts away the calculation of the best tip height based on
the finalized block height and the best non-finalized chain's tip.
* Add `best_tip_height` field to `StateService`
The receiver endpoint is currently ignored.
* Return receiver endpoint from service constructor
Make it available so that the best tip height can be watched.
* Update finalized height after finalizing blocks
After blocks from the queue are finalized and committed to disk, update
the finalized block height.
* Update best non-finalized height after validation
Update the value of the best non-finalized chain tip block height after
a new block is committed to the non-finalized state.
* Update finalized height after loading from disk
When `FinalizedState` is first created, it loads the state from
persistent storage, and the finalized tip height is updated. Therefore,
the `best_tip_height` must be notified of the initial value.
* Update the finalized height on checkpoint commit
When a checkpointed block is commited, it bypasses the non-finalized
state, so there's an extra place where the finalized height has to be
updated.
* Add `best_tip_height` to `Handshake` service
It can be configured using the `Builder::with_best_tip_height`. It's
currently not used, but it will be used to determine if a connection to
a remote peer should be rejected or not based on that peer's protocol
version.
* Require best tip height to init. `zebra_network`
Without it the handshake service can't properly enforce the minimum
network protocol version from peers. Zebrad obtains the best tip height
endpoint from `zebra_state`, and the test vectors simply use a dummy
endpoint that's fixed at the genesis height.
* Pass `best_tip_height` to proto. ver. negotiation
The protocol version negotiation code will reject connections to peers
if they are using an old protocol version. An old version is determined
based on the current known best chain tip height.
* Handle an optional height in `Version`
Fallback to the genesis height in `None` is specified.
* Reject connections to peers on old proto. versions
Avoid connecting to peers that are on protocol versions that don't
recognize a network update.
* Document why peers on old versions are rejected
Describe why it's a security issue above the check.
* Test if `BestTipHeight` starts with `None`
Check if initially there is no best tip height.
* Test if best tip height is max. of latest values
After applying a list of random updates where each one either sets the
finalized height or the non-finalized height, check that the best tip
height is the maximum of the most recently set finalized height and the
most recently set non-finalized height.
* Add `queue_and_commit_finalized` method
A small refactor to make testing easier. The handling of requests for
committing non-finalized and finalized blocks is now more consistent.
* Add `assert_block_can_be_validated` helper
Refactor to move into a separate method some assertions that are done
before a block is validated. This is to allow moving these assertions
more easily to simplify testing.
* Remove redundant PoW block assertion
It's also checked in
`zebra_state::service::check::block_is_contextually_valid`, and it was
getting in the way of tests that received a gossiped block before
finalizing enough blocks.
* Create a test strategy for test vector chain
Splits a chain loaded from the test vectors in two parts, containing the
blocks to finalize and the blocks to keep in the non-finalized state.
* Test committing blocks update best tip height
Create a mock blockchain state, with a chain of finalized blocks and a
chain of non-finalized blocks. Commit all the blocks appropriately, and
verify that the best tip height is updated.
Co-authored-by: teor <teor@riseup.net>
* Use the block verifier and non-finalized state in the cached state tests
This substantially increases test coverage.
Previously, the cached state tests were configured with
`checkpoint_sync = true`, which only uses the checkpoint
verifier and the finalized state.
* Log the source of blocks in commit_finalized_direct
This lets us check that we're actually testing the non-finalized state
and block verifier in the cached state tests.
It also improves diagnostics for state errors.
* Fail cached state tests if they're using incorrect heights or configs
This makes sure that the cached state tests actually test the transition
from checkpoint to block verification, and the non-finalized state.
* Update versions for zebra v1.0.0-alpha.12 release
* Update Cargo.lock
* Update release checklist with latest version changes to help keep track for future releases
* Remove reference to the fact that tower-fallback was not updated
* add legacy chain check and tests
* improve has_network_upgrade check
* add docs to legacy_chain_check()
* change arbitrary module structure
* change the panic message
* move legacy chain acceptance into existing tests
* use a reduced_branch_id_strategy()
* add docs to strategy function
* add argument to check for legacy chain into sync_until()
* Disable IPv6 tests when $ZEBRA_SKIP_IPV6_TESTS is set
This allows users to disable IPv6 tests in environments where IPv6 is not
configured.
* Add network test env var constants
* Replace env strings with constants
fastmod '"ZEBRA_SKIP_NETWORK_TESTS"' zebra_test::net::ZEBRA_SKIP_NETWORK_TESTS
fastmod '"ZEBRA_SKIP_IPV6_TESTS"' zebra_test::net::ZEBRA_SKIP_IPV6_TESTS
* Add functions to skip network tests
* Replace test network env var checks with test function
fastmod --fixed-strings 'env::var_os(zebra_test::net::ZEBRA_SKIP_NETWORK_TESTS).is_some()' 'zebra_test::net::zebra_skip_network_tests()'
fastmod --fixed-strings 'env::var_os(zebra_test::net::ZEBRA_SKIP_IPV6_TESTS).is_some()' 'zebra_test::net::zebra_skip_ipv6_tests()'
* Remove redundant logging and use statements
* Gossip dynamically allocated listener ports to peers
Previously, Zebra would either gossip port `0`, which is invalid, or skip
gossiping its own dynamically allocated listener port.
* Improve "no configured peers" warning
And downgrade from error to warning, because inbound-only nodes are a
valid use case.
* Move random_known_port to zebra-test
* Add tests for dynamic local listener ports and the AddressBook
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
* Always send our local listener with the latest time
Previously, whenever there was an inbound request for peers, we would
clone the address book and update it with the local listener.
This had two impacts:
- the listener could conflict with an existing entry,
rather than unconditionally replacing it, and
- the listener was briefly included in the address book metrics.
As a side-effect, this change also makes sanitization slightly faster,
because it avoids some useless peer filtering and sorting.
* Skip listeners that are not valid for outbound connections
* Filter sanitized addresses Zebra based on address state
This fix correctly prevents Zebra gossiping client addresses to peers,
but still keeps the client in the address book to avoid reconnections.
* Add a full set of DateTime32 and Duration32 calculation methods
* Refactor sanitize to use the new DateTime32/Duration32 methods
* Security: Use canonical SocketAddrs to avoid duplicate connections
If we allow multiple variants for each peer address, we can make multiple
connections to that peer.
Also make sure sanitized MetaAddrs are valid for outbound connections.
* Test that address books contain the local listener address
Co-authored-by: Janito Vaqueiro Ferreira Filho <janito.vff@gmail.com>
- Add a custom semver match for `zebrad` versions
- Prefer "line contains string" matches, so tests ignore minor changes
- Escape regex meta-characters when a literal string match is intended
- Rename test functions so they are more precise
- Rewrite match internals to remove duplicate code and enable custom matches
- Document match functions
* implement and test a rate limit in `request_genesis()`
* add `request_genesis_is_rate_limited` test to sync
* add ensure_timeouts constraint for GENESIS_TIMEOUT_RETRY
* Suppress expected warning logs in zebrad tests
Co-authored-by: teor <teor@riseup.net>
* Standardise lints across Zebra crates, and add missing docs
The only remaining module with missing docs is `zebra_test::command`
* Todo -> TODO
* Clarify what a transcript ErrorChecker does
Also change `Error` -> `BoxError`
* TransError -> ExpectedTranscriptError
* Output Descriptions -> Output descriptions
When peers ask for peer addresses, add our local listener address to the
set of addresses, sanitize, then truncate. Sanitize shuffles addresses,
so if there are lots of addresses in the address book, our address will
only be sent to some peers.
* Use the git version + new commit count + hash for the app version
This helps diagnose bugs in versions of Zebra built from git branches,
rather than git version tags.
* Fill in assert
* Also log semver string
* Fix syntax
* Handle vergen using the cargo package version or raw git tag
* s/Semver/SemVer/
Co-authored-by: Deirdre Connolly <durumcrustulum@gmail.com>
Enable builds where:
* there is no google cloud git commit env var, and
* there is no `.git` directory.
By making all `vergen` env vars optional, and skipping any env vars that
don't exist.
* build(deps): bump vergen from 3.2.0 to 5.1.1
* fix hardcoded version for Tracing struct
* add additional metadata
* remove extra allocations for metadata
* Remove zebrad code version from release checklist
The zebrad code automatically uses the crate version now.
* Sort panic metadata into rough categories
Co-authored-by: teor <teor@riseup.net>
Zebra's latest alpha checkpoints on Canopy activation, continues our work on NU5, and fixes a security issue.
Some notable changes include:
## Added
- Log address book metrics when PeerSet or CandidateSet don't have many peers (#1906)
- Document test coverage workflow (#1919)
- Add a final job to CI, so we can easily require all the CI jobs to pass (#1927)
## Changed
- Zebra has moved its mandatory checkpoint from Sapling to Canopy (#1898, #1926)
- This is a breaking change for users that depend on the exact height of the mandatory checkpoint.
## Fixed
- tower-batch: wake waiting workers on close to avoid hangs (#1908)
- Assert that pre-Canopy blocks use checkpointing (#1909)
- Fix CI disk space usage by disabling incremental compilation in coverage builds (#1923)
## Security
- Stop relying on unchecked length fields when preallocating vectors (#1925)
`node2.is_running()` can return `true` on Windows, even though `node2`
has logged a panic. This cleanup code only runs if `node2` fails to panic
and exit as expected. So it's ok for us to skip it.
See #1781 for details.
On Windows, if a process is killed after it is dead, it returns `true`
for `was_killed`. Instead, check if the process is running before killing
it.
Also make the section where processes are running as short as possible,
and include context for both processes in every error.
Log a "Trying..." message before each listener opens, to see if the
delay is inside Zebra, or in the test harness or OS.
Also report the configured and actual ports where possible, for better
diagnostics.
* Increase the conflict acceptance test launch delay
Also rename the tests - the listener is for the Zcash protocol,
but the state, metrics, and tracing are Zebra-specific.
Co-authored-by: Alfredo Garcia <oxarbitrage@gmail.com>
This change encodes a bunch of invariants in the type system,
and adds explicit failure states for:
* a closed oneshot,
* bugs in the initialization code.
Use `ServiceExt::oneshot` to perform state requests.
Explain that `ServiceExt::call_all` calls `poll_ready` internally.
Document a state service invariant imposed by `ServiceExt::call_all`.
* add hint for port error
* add issue filter for port panic
* add lock file hint
* add metrics endpoint port conflict hint
* add hint for tracing endpoint port conflict
* add acceptance test for resource conflics
* Split out common conflict test code into a function
* Add state, metrics, and tracing conflict tests
* Add a full set of stderr acceptance test functions
This change makes the stdout and stderr acceptance test interfaces
identical.
* move Zcash listener opening
* add todo about hint for disk full
* add constant for lock file
* match path in state cache
* don't match windows cache path
* Use Display for state path logs
Avoids weird escaping on Windows when using Debug
* Add Windows conflict error messages
* Turn PORT_IN_USE_ERROR into a regex
And add another alternative Windows-specific port error
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Jane Lusby <jane@zfnd.org>
* Bump versions where appropriate
Tested with cargo install --locked --path etc
* Remove fixed panics from 'Known Issues'
* Change to alpha release series in the README
Co-authored-by: teor <teor@riseup.net>
The clippy unknown lints attribute was deprecated in
nightly in rust-lang/rust#80524. The old lint name now produces a
warning.
Since we're using `allow(unknown_lints)` to suppress warnings, we need to
add the canonical name, so we can continue to build without warnings on
nightly.
But we also need to keep the old name, so we can continue to build
without warnings on stable.
And therefore, we also need to disable the "removed lints" warning,
otherwise we'll get warnings about the old name on nightly.
We'll need to keep this transitional clippy config until rustc 1.51 is
stable.
* Stop failing acceptance tests if their directories already exist
* Add an immutable config writing helper
and use it in the cached sapling acceptance tests.
Also:
* consistently create missing config and state directories
* refactor the common config writing code into a separate function
* only ignore NotFound errors in replace_config
* enforce config immutability using the type system
This timeout stops the sync service hanging when it is missing required
blocks, but the lookahead queue is full of dependent verify tasks, so the
missing blocks never get downloaded.
Check misconfigured ephemeral doesn't create a state dir
Add extra misconfigured `zebrad` ephemeral mode checks:
* doesn't create a state directory
* doesn't create unexpected files or directories in the working directory
Check ephemeral doesn't delete an existing state directory
Refactor all the ephemeral configs and checks into a single test
function.
Also:
* cleanup acceptance tests using utility functions
* make some checks consistent between tests
* make error messages consistent
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* Add the configured network to error reports
* Log the configured network at error level
* Create the global span immediately after activating tracing
And leak the span guard, so the span is always active.
* Include panic metadata in the report and URL
* Use `Main` and `Test` in the global span
`net=Mainnet` is a bit redundant
When `cargo run` is run in the workspace directory, it can see two
executables:
- `zebrad`
- `zebra_checkpoints`
Adding `default-run = "zebrad"` to `zebrad/Cargo.toml` makes the
workspace run `zebrad` by default. (Even though it's redundant for the
`zebrad` crate itself.)
* Rewrite GetData handling to match the zcashd implementation
`zcashd` silently ignores missing blocks, but sends found transactions
followed by a `NotFound` message:
e7b425298f/src/main.cpp (L5497)
This is significantly different to the behaviour expected by the old
Zebra connection state machine, which expected `NotFound` for blocks.
Also change Zebra's GetData responses to peer request so they ignore
missing blocks.
* Stop hanging on incomplete transaction or block responses
Instead, if the peer sends an unexpected block, unexpected transaction,
or NotFound message:
1. end the request, and return a partial response containing any items
that were successfully received
2. if none of the expected blocks or transactions were received, return
an error, and close the connection
In our README, we tell users to ignore these errors, so we should also
disable the issue URL.
Also include the hash in the error. (We don't want the span active for
all messages, we just want the hash in the error.)
This change avoids errors when tests are cancelled and re-run within a
short period of time, for example, using `cargo watch`.
It introduces a slight risk of port conflicts between the endpoint tests,
and with (ephemeral) ports used by other services. The risk of conflicts
across 2 tests is very low, and tests should be run in an isolated
environment on busy servers.
vergen's implementation of REBUILD_ON_HEAD_CHANGE assumes that the .git
directory is in the crate root, but Zebra uses a workspace.
Temporary fix for rustyhorde/vergen#21.
Because the new version of the prometheus exporter launches its own
single-threaded runtime on a dedicated worker thread, there's no need
for the tokio and hyper versions it uses internally to align with the
versions used in other crates. So we don't need to use our fork with
tokio 0.3, and can just use the published alpha. Advancing to a later
alpha may fix the missing-metrics issues.
As we approach our alpha release we've decided we want to plan ahead for the user bug reports we will eventually receive. One of the bigger issues we foresee is determining exactly what version of the software users are running, and particularly how easy it may or may not be for users to accidentally discard this information when reporting bugs.
To defend against this, we've decided to include the exact git sha for any given build in the compiled artifact. This information will then be re-exported as a span early in the application startup process, so that all logs and error messages should include the sha as their very first span. We've also added this sha as issue metadata for `color-eyre`'s github issue url auto generation feature, which should make sure that the sha is easily available in bug reports we receive, even in the absence of logs.
Co-authored-by: teor <teor@riseup.net>
* implement inbound `FindBlocks`
* Handle inbound peer FindHeaders requests
* handle request before having any chain tip
* Split `find_chain_hashes` into smaller functions
Add a `max_len` argument to support `FindHeaders` requests.
Rewrite the hash collection code to use heights, so we can handle the
`stop` hash and "no intersection" cases correctly.
* Split state height functions into "any chain" and "best chain"
* Rename the best chain block method to `best_block`
* Move fmt utilities to zebra_chain::fmt
* Summarise Debug for some Message variants
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
This provides useful and not too noisy output at INFO level. We do an
info-level message on every block commit instead of trying to do one
message every N blocks, because this is useful both for initial block
sync as well as continuous state updates on new blocks.
The metrics code becomes much simpler because the current version of the
metrics crate builds its own single-threaded runtime on a dedicated worker
thread, so no dependency on the main Zebra Tokio runtime is required.
This change is mostly mechanical, with the exception of the changes to the
`tower-batch` middleware. This middleware was adapted from `tower::buffer`,
and the `tower::buffer` code was changed to implement its own bounded queue,
because Tokio 0.3 removed the `mpsc::Sender::poll_send` method. See
ddc64e8d4d
for more context on the Tower changes. To match Tower as closely as possible
in order to be able to upstream `tower-batch`, those changes are copied from
`tower::Buffer` to `tower-batch`.
## Motivation
Prior to this PR we've been using `sled` as our database for storing persistent chain data on the disk between boots. We picked sled over rocksdb to minimize our c++ dependencies despite it being a less mature codebase. The theory was if it worked well enough we'd prefer to have a pure rust codebase, but if we ever ran into problems we knew we could easily swap it out with rocksdb.
Well, we ran into problems. Sled's memory usage was particularly high, and it seemed to be leaking memory. On top of all that, the performance for writes was pretty poor, causing us to become bottle-necked on sled instead of the network.
## Solution
This PR replaces `sled` with `rocksdb`. We've seen a 10x improvement in memory usage out of the box, no more leaking, and much better write performance. With this change writing chain data to disk is no longer a limiting factor in how quickly we can sync the chain.
The code in this pull request has:
- [x] Documentation Comments
- [x] Unit Tests and Property Tests
## Review
@hdevalence
This helps prevent overloading the network with too many concurrent
block requests. On a fast network, we're likely to still have enough
room to saturate our bandwidth. In the worst case, with 2MB blocks,
downloading 50 blocks concurrently is 100MB of queued downloads. If we
need to download this in 20 seconds to avoid peer connection timeouts,
the implied worst-case minimum speed is 5MB/s. In practice, this
minimum speed will likely be much lower.
This reverts commit 656bd24ba7.
The Hedge middleware keeps a pair of histograms, writing into one in the
current time interval and reading from the previous time interval's
data. This means that the reverted change resulted in doubling all
block downloads until after at least the second measurement interval
(which means that the time measurements are also incorrect, as they're
operating under double the network load...)
* Use the default memory limit in the acceptance tests
PR #1233 changed the default `memory_cache_bytes`, but left the
acceptance tests with their old value.
Sets the default value to the previous lookahead limit. My testing on
mainnet suggested that the newly lower value (changed when the
checkpoint frequency was decreased) is low enough to cause stalls, even
when using hedged requests.
Remove the minimum data points from the syncer hedge configuragtion.
When there are no data points, hedge sends the second request
immediately.
Where there are less than 1/(1-latency_percentile) data points (20),
hedge delays the second request by the highest recent download time.
This change should improve genesis and post-restart sync latency.
* Run large checkpoint sync tests in CI
* Improve test child output match error context
* Add a debug_stop_at_height config
* Use stop at height in acceptance tests
And add some restart acceptance tests, to make sure the stop at
height feature works correctly.
We should error if we notice that we're attempting to download the same
blocks multiple times, because that indicates that peers reported bad
information to us, or we got confused trying to interpret their
responses.
The original sync algorithm split the sync process into two phases, one
that obtained prospective chain tips, and another that attempted to
extend those chain tips as far as possible until encountering an error
(at which point the prospective state is discarded and the process
restarts).
Because a previous implementation of this algorithm didn't properly
enforce linkage between segments of the chain while extending tips,
sometimes it would get confused and fail to discard responses that did
not extend a tip. To mitigate this, a check against the state was
added. However, this check can cause stalls while checkpointing,
because when a checkpoint is reached we may suddenly need to commit
thousands of blocks to the state. Because the sync algorithm now has a
a `CheckedTip` structure that ensures that a new segment of hashes
actually extends an existing one, we don't need to check against the
state while extending a tip, because we don't get confused while
interpreting responses.
This change results in significantly smoother progress on mainnet.
There's no reason to return a pre-Buffer'd service (there's no need for
internal access to the state service, as in zebra-network), but wrapping
it internally removes control of the buffer size from the caller.
The timeout behavior in zebra-network is an implementation detail, not a
feature of the public API. So it shouldn't be mentioned in the doc
comments -- if we want timeout behavior, we have to layer it ourselves.
Using the cancel_handles, we can deduplicate requests. This is
important to do, because otherwise when we insert the second cancel
handle, we'd drop the first one, cancelling an existing task for no
reason.
The hedge middleware implements hedged requests, as described in _The
Tail At Scale_. The idea is that we auto-tune our retry logic according
to the actual network conditions, pre-emptively retrying requests that
exceed some latency percentile. This would hopefully solve the problem
where our timeouts are too long on mainnet and too slow on testnet.
Try to use the better cancellation logic to revert to previous sync
algorithm. As designed, the sync algorithm is supposed to proceed by
downloading state prospectively and handle errors by flushing the
pipeline and starting over. This hasn't worked well, because we didn't
previously cancel tasks properly. Now that we can, try to use something
in the spirit of the original sync algorithm.
This makes two changes relative to the existing download code:
1. It uses a oneshot to attempt to cancel the download task after it
has started;
2. It encapsulates the download creation and cancellation logic into a
Downloads struct.
* Reverse displayed endianness of transaction and block hashes
* fix zebra-checkpoints utility for new hash order
* Stop using "zebrad revhex" in zebrad-hash-lookup
* Rebuild checkpoint lists in new hash order
This change also adds additional checkpoints to the end of each list.
* Replace TransactionHash with transaction::Hash
This change should have been made in #905, but we missed Debug impls
and some docs.
Co-authored-by: Ramana Venkata <vramana@users.noreply.github.com>
Co-authored-by: teor <teor@riseup.net>
This reduces the API surface to the minimum required for functionality,
and cleans up module documentation. The stub mempool module is deleted
entirely, since it will need to be redone later anyways.
* implement most of the chain functions
* implement fork
* fix outpoint handling in Chain struct
* update expect for work
* split utxo into two sets
* update the Chain definition
* remove allow attribute in zebra-state/lib.rs
* merge ChainSet type into MemoryState
* Add error messages to asserts
* export proptest impls for use in downstream crates
* add testjob for disabled feature in zebra-chain
* try to fix github actions syntax
* add module doc comment
* update RFC for utxos
* add missing header
* working proptest for Chain
* propagate back results over channel
* Start updating RFC to match changes
* implement queued block pruning
* and now it syncs wooo!
* remove empty modules
* setup config for proptests
* re-enable missing_docs lint
* update RFC to match changes in impl
* add documentation
* use more explicit variable names
The Inbound service only needs the network setup for some requests, but
it can service other requests without it. Making it return
Poll::Pending until the network setup finishes means that initial
network connections may view the Inbound service as overloaded and
attempt to load-shed.
The original version of this commit ran into
https://github.com/rust-lang/rust/issues/64552
again. Thanks to @yaahc for suggesting a workaround (using futures combinators
to avoid writing an async block).
Remove the seed command entirely, and make the behavior it provided
(responding to `Request::Peers`) part of the ordinary functioning of the
start command.
The new `Inbound` service should be expanded to handle all request
types.
* add test for first checkpoint sync
Prior this this change we've not had any tests that verify our sync /
network logic is well behaved. This PR cleans up the test helper code to
make error reports more consistent and uses this cleaned up API to
implement a checkpoint sync test which runs zebrad until it reads the
first checkpoint event from stdout.
Co-authored-by: teor <teor@riseup.net>
* move include out of unix cfg
Co-authored-by: teor <teor@riseup.net>
* increase the EWMA default and decay
* increase the block download retries
* increase the request and block download timeouts
* increase the sync timeout
* add metrics and tracking endpoint tests
* test endpoints more
* add change filter test for tracing
* add await to post
* separate metrics and tracing tests
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: teor <teor@riseup.net>
* Ignore sync errors when the block is already verified
If we get an error for a block that is already in our state, we don't
need to restart the sync. It was probably a duplicate download.
Also:
Process any ready tasks before reset, so the logs and metrics are
up to date. (But ignore the errors, because we're about to reset.)
Improve sync logging and metrics during the download and verify task.
* Remove duplicate hashes in logs
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* Log the sync hash span at warn level
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* Remove in-memory state service
* make the config compatible with toml again
* checkpoint commit to see how much I still have to revert
* back to the starting point...
* remove unused dependency
* reorganize error handling a bit
* need to make a new color-eyre release now
* reorder again because I have problems
* remove unnecessary helpers
* revert changes to config loading
* add back missing space
* Switch to released color-eyre version
* add back missing newline again...
* improve error message on unix when terminated by signal
* add context to last few asserts in acceptance tests
* instrument some of the helpers
* remove accidental extra space
* try to make this compile on windows
* reorg platform specific code
* hide on_disk module and fix broken link
* add CheckpointList::new_up_to(limit: NetworkUpgrade)
* if checkpoint_sync is false, limit checkpoints to Sapling
* update tests for CheckpointList and chain::init
This is the first in a sequence of changes that change the block:: items
to not include Block as a prefix in their name, in accordance with the
Rust API guidelines.
* add valid generated config test
* change to pathbuf
* use -c to make sure we are using the generated file
* add and use a ZebraTestDir type
* change approach to generate tempdir in top of each test
* pass tempdir to test_cmd and set current dir to it
* add and use a `generated_config_path` variable in tests
* checkpoint: reject older of duplicate verification requests.
If we get a duplicate block verification request, we should drop the older one
in favor of the newer one, because the older request is likely to have been
canceled. Previously, this code would accept up to four duplicate verification
requests, then fail all subsequent ones.
* sync: add a timeout layer to block requests.
Note that if this timeout is too short, we'll bring down the peer set in a
retry storm.
* sync: restart syncing on error
Restart the syncing process when an error occurs, rather than ignoring it.
Restarting means we discard all tips and start over with a new block locator,
so we can have another chance to "unstuck" ourselves.
* sync: additional debug info
* sync: handle lookahead limit correctly.
Instead of extracting all the completed task results, the previous code pulled
results out until there were fewer tasks than the lookahead limit, then
stopped. This meant that completed tasks could be left until the limit was
exceeded again. Instead, extract all completed results, and use the number of
pending tasks to decide whether to extend the tip or wait for blocks to finish.
* network: add debug instrumentation to retry policy
* sync: instrument the spawned task
* sync: streamline ObtainTips/ExtendTips logic & tracing
This change does three things:
1. It aligns the implementation of ObtainTips and ExtendTips so that they use
the same deduplication method. This means that when debugging we only have one
deduplication algorithm to focus on.
2. It streamlines the tracing output to not include information already
included in spans. Both obtain_tips and extend_tips have their own spans
attached to the events, so it's not necessary to add Scope: prefixes in
messages.
3. It changes the messages to be focused on reporting the actual
events rather than the interpretation of the events (e.g., "got genesis hash in
response" rather than "peer could not extend tip"). The motivation for this
change is that when debugging, the interpretation of events is already known to
be incorrect, in the sense that the mental model of the code (no bug) does not
match its behavior (has bug), so presenting minimally-interpreted events forces
interpretation relative to the actual code.
* sync: hack to work around zcashd behavior
* sync: localize debug statement in extend_tips
* sync: change algorithm to define tips as pairs of hashes.
This is different enough from the existing description that its comments no
longer apply, so I removed them. A further chunk of work is to change the sync
RFC to document this algorithm.
* sync: reduce block timeout
* state: add resource limits for sled
Closes#888
* sync: add a restart timeout constant
* sync: de-pub constants
* network: move gossiped peer selection logic into address book.
* network: return BoxService from init.
* zebrad: add note on why we truncate thegossiped peer list
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* Remove unused .rustfmt.toml
Many of these options are never actually loaded by our CI because of a channel
mismatch, where they're not applied on stable but only on nightly (see the logs
from a rustfmt job). This means that we can get different settings when
running `cargo fmt` on the nightly and stable channels, which was causing a CI
failure on this PR. Reverting back to the default rustfmt settings avoids this
problem and keeps us in line with upstream rustfmt. There's no loss to us
since we were using the defaults anyways.
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* fix: Handle known ObtainTips correctly
enumerate never returns a value beyond the end of the vector.
* fix: Ignore known tips in ExtendTips
Some peers send us known tips when we try to extend.
* fix: Ignore known hashes when downloading
Despite all our other checks, we still end up downloading some hashes
multiple times.
* fix: Increase the number of retries
The old sync code relied on duplicate block fetches to make progress,
but the last few commits have removed some of those duplicates.
Instead, just retry the fetches that fail.
* fix: Tweak comments
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* fix: Cleanup the state_contains interface in Sync
* Fix brackets
Oops
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
* make sure no info is printed in non server tests
* check exact full output for validity instead of log msgs
* add end of output character to version regex
* use coercions, use equality operator
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
Co-authored-by: Jane Lusby <jlusby42@gmail.com>
The old sync code relied on duplicate block fetches to make progress,
but the last few commits have removed some of those duplicates.
Instead, just retry the fetches that fail.
We have a counter for pending "download and verify" futures. But these
futures are spawned, so they can complete in any order. They can also
complete before we receive their results.
* Split tracing component code into modules.
* Repatriate Tracing and simplify config handling.
We upstreamed our Tracing component, expecting not to have to exert fine
control over the tracing settings. But this turned out not to be the case, and
now that we want to do other things (flamegraphs, journalctl, opentelemetry,
etc), we end up with really awkward code (as in the current flamegraph
handling).
This also makes use of the changes to `init()` to load the config early to pass
configuration data into the components, which avoids the need for the
refactoring in #775.
Finally, we restore support for the `-v` flag when the filter is unset. Closes#831.
* Disable tracing and metrics endpoints by default.
Closes#660.
* Switch back to upstream Abscissa.
* Integrate flamegraph support into the new Tracing component.
* Pass -v in acceptance tests to get info-level output.
* Clean up acceptance test code.
* Setup tracing-flame for use profiling zebrad
* start work on conditional flamegraph generation
* review time!
* update comments
* Update Cargo.toml
* disable default features for inferno
* reorganize
* missing one trait
* Apply suggestions from code review
* graceful shutdown!
* remove special case handling on ctrlc for cleanup
* rename signal fn to better represent its responsibility
* remove unused global hook for flushing flamegraph
* move tracing logic to the right file
* just copy linkerd's signal handling logic
* update book
* make zebrad app drop on shutdown normally
* Update zebrad/src/components/tokio.rs
Co-authored-by: teor <teor@riseup.net>
* Update zebrad/src/application.rs
Co-authored-by: teor <teor@riseup.net>
* Apply suggestions from code review
Co-authored-by: teor <teor@riseup.net>
* cleanup a little
* ooh yea there's an API for that
* setup env-filter for backup subscriber
* document env filter
* document return codes
* forgot to save
* Update book/src/applications/zebrad.md
Co-authored-by: teor <teor@riseup.net>
Co-authored-by: teor <teor@riseup.net>
* Load tracing filter only from config and simplify logic.
* Configure the state storage in the config, not an environment variable.
This also changes the config so that the path is always set rather than being
optional, because Zebra always needs a place to store its config.
* add zebrad acceptance tests
* add custom command test helpers that work with kill
* add and use info event for start and seed commands
* combine conflicting tests into one test case
Co-authored-by: Jane Lusby <jane@zfnd.org>
We get the injected TokioComponent dependency before the config is
loaded, so we can't use it to open the endpoints.
And we can't define after_config, because we use derive(Component).
So we work around these issues by opening the endpoints manually,
from the application's after_config.
* make zebra-checkpoints
* fix LOOKAHEAD_LIMIT scope
* add a default cli path
* change doc usage text
* add tracing
* move MAX_CHECKPOINT_HEIGHT_GAP to zebra-consensus
* do byte_reverse_hex in a map
This seems like a better design on principle but also appears to give a much
nicer sawtooth pattern of queued blocks in the checkpointer and a much smoother
pattern of block requests.
We had a brief discussion on discord and it seemed like we had consensus on the
following versioning policy:
* zebrad: match major version to NU version, so we will start by releasing
zebrad 3.0.0;
* zebra-* libraries: start by matching zebrad's version, then increment major
versions of each library as we need to make breaking changes (potentially
faster than the zebrad version, always respecting semver but making no
guarantees about the longevity of major releases).
This commit sets all of the crate versions to 3.0.0-alpha.0 -- the -alpha.0
marks it as a prerelease not subject to perfect adherence to compatibility
guarantees.
Conrado Gouvea
teor
Janito Vaqueiro Ferreira Filho
Alfredo Garcia
dependabot[bot]
Marek
Deirdre Connolly
Deirdre Connolly
Dimitris Apostolou
Pili Guerra
Conrado P. L. Gouvea
Kirill Fomichev
Jack Grigg
Jane Lusby
Jane Lusby
Henry de Valence
Ramana Venkata