feat(store/v2): snapshot manager #18458
Conversation
WalkthroughThe changes involve a significant overhaul of the snapshot management system, introducing new interfaces and methods for snapshot creation and restoration, as well as pruning and storage management. The Changes
Related issues
TipsChat with CodeRabbit Bot (
|
|
@cool-develope your pull request is missing a changelog! |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Review Status
Actionable comments generated: 4
Configuration used: CodeRabbit UI
Files ignored due to filter (1)
- store/go.mod
Files selected for processing (9)
- store/commitment/iavl/tree.go (2 hunks)
- store/commitment/iavl/tree_test.go (2 hunks)
- store/pruning/manager_test.go (1 hunks)
- store/root/store_test.go (1 hunks)
- store/snapshots/helpers_test.go (6 hunks)
- store/snapshots/manager.go (6 hunks)
- store/snapshots/manager_test.go (8 hunks)
- store/snapshots/types/snapshotter.go (1 hunks)
- store/storage/sqlite/db.go (4 hunks)
Additional comments: 42
store/pruning/manager_test.go (2)
34-34: The
NewIavlTreefunction is called with an empty string argument beforeiavl.DefaultConfig(). This empty string argument is likely a placeholder for a parameter that was either removed or refactored. Verify that theNewIavlTreefunction's signature and the expected arguments are correct and that this change does not introduce any unintended behavior.36-36: The
Managerstruct is being initialized withnoopLog,ss, andscas arguments. Ensure that theNewManagerfunction's signature has been updated to accept these arguments and that theManagerstruct's fields are correctly assigned within theNewManagerfunction.store/root/store_test.go (1)
- 30-36: The
SetupTestfunction has been updated to include thedefaultStoreKeyparameter when initializing thescvariable. This change should be verified to ensure that it aligns with the intended use ofdefaultStoreKeyand that it does not introduce any unintended side effects in the test suite or theIavlTreeinitialization.Overall, the changes in the test suite seem to be in response to the updated interfaces and functionality of the store component. It is important to ensure that all tests pass and that the new functionality works as expected, especially with regards to the snapshot manager feature and the backward compatibility with the old snapshot format.
store/snapshots/types/snapshotter.go (1)
- 2-23: The interfaces
CommitSnapshotterandStorageSnapshotterhave been clearly defined with their respective responsibilities. TheSnapshotmethod inCommitSnapshotternow correctly takes aversionargument, which is more generic and can be applied to different types of stores, not just those that use a height-based versioning system. TheRestoremethod inCommitSnapshotteralso appropriately takes a channel for storage key-value pairs, which allows for asynchronous and potentially concurrent restoration of state. The removal ofPruneSnapshotHeightandSetSnapshotIntervalfunctions suggests that snapshot management is being refactored to a different part of the system or being handled differently, which should be verified for compatibility with existing systems.25-:
TheExtensionSnapshotterinterface remains unchanged in this hunk. It's important to ensure that the unchanged methods still align with the new snapshotting strategy and that any implementations of this interface are updated accordingly if needed.store/commitment/iavl/tree_test.go (6)
3-15: The addition of new imports suggests that new dependencies are being used in the test file. Ensure that these dependencies are properly managed and that the import paths are correct.
17-21: The
generateTreefunction has been modified to accept astoreKeyparameter. This change should be reflected wherever the function is called to ensure compatibility.98-156: The new test function
TestSnapshotteris added to test the snapshot creation and restoration functionality. It is important to ensure that this test covers all relevant cases and that the snapshot and restore operations are working as expected. The use of channels and goroutines should be carefully reviewed to avoid data races.131-131: The
generateTreefunction is called with an empty string for thestoreKeyparameter. This might be intentional for testing purposes, but it's important to ensure that this is the expected behavior and that thestoreKeyis not required to be non-empty for theSnapshotandRestoremethods to function correctly.132-142: The use of a goroutine to write to the
chunkschannel and the subsequent reading from this channel instreamReadercould potentially lead to data races if not properly synchronized. Ensure that the channel is closed after all writes are completed to prevent deadlocks or reading from a closed channel.147-155: The
Restoremethod is tested here. It is crucial to ensure that theRestoremethod's error handling is robust, especially considering that it reads from a channel which could be closed or could contain unexpected data.store/storage/sqlite/db.go (2)
11-15: The import block is clean and well-organized.
44-51: The
Databasestruct correctly implements thestore.VersionedDatabaseandsnapshottypes.StorageSnapshotterinterfaces, as indicated by thevar _ Interface = (*Struct)(nil)pattern. This is a common Go idiom to ensure at compile time that the struct satisfies the interface.store/snapshots/manager_test.go (8)
17-25: The test setup for
TestManager_Listlooks correct, and the assertions are appropriate for the functionality being tested. The test checks both the normal operation and the behavior when the manager is busy.40-46: In
TestManager_LoadChunk, the test setup and assertions are correct. It tests for the presence of an existing chunk, the absence of a non-existing chunk, and the behavior under a busy manager.66-78:
TestManager_Takecorrectly sets up amockCommitSnapshotterwith predefined items and tests snapshot creation. The test also checks for error conditions such as creating a snapshot at a lower height than the latest and while another snapshot is being created.83-96: The test checks for successful snapshot creation at a higher height and validates the snapshot's properties. The use of
commitSnapshotter.SnapshotFormat()to set the format in the expected snapshot is a good practice as it ensures consistency with the actual implementation.111-117:
TestManager_Prunetests the pruning functionality and checks for errors when a snapshot is being taken. The test setup and assertions are correct.129-137:
TestManager_Restoresets up the necessary mock snapshotter and extension snapshotter. It tests various error conditions for the restore operation, such as invalid format, no chunks, and chunk/hash mismatches.181-186: The test checks that while a restore is in progress, other operations such as creating a snapshot or pruning should fail. This is an important concurrency test to ensure that the manager's state is correctly handled.
236-245:
TestManager_TakeErrorsets up amockErrorCommitSnapshotterto simulate an error during snapshot creation. The test correctly asserts that an error is returned.store/snapshots/manager.go (5)
35-46: The
Managerstruct has been updated to includecommitSnapshotterandstorageSnapshotterfields, and themultistorefield has been removed. This change aligns with the new snapshot interfaces and should be checked for compatibility with existing code that may have used themultistorefield.65-73: Introduction of
defaultStorageChannelBufferSizeas a global variable. This should be reviewed to ensure that the buffer size is appropriate for the expected workload and that it does not introduce memory issues.77-90: The
NewManagerfunction signature has been updated to acceptcommitSnapshotterandstorageSnapshotterparameters. This change requires verification that all calls toNewManagerhave been updated accordingly.204-210: The
createSnapshotfunction now uses thecommitSnapshotterto take a snapshot at a given height. It is important to ensure that thecommitSnapshotteris properly initialized before this function is called to avoid runtime errors.366-385: > Note: This review was outside of the patch, so it was mapped to the patch with the greatest overlap. Original lines [366-419]
The
doRestoreSnapshotfunction has been updated to use thestorageSnapshotterfor restoring the storage state. It is crucial to ensure that thestorageSnapshotteris correctly handling the key-value pairs and that the error handling is robust, especially in the go-routine that performs the restoration.The overall changes to the
Managerstruct and related functions appear to be part of a significant refactor to support a new snapshotting mechanism. It is essential to thoroughly test these changes to ensure that the snapshot creation and restoration processes work as expected and that the new interfaces are correctly implemented. Additionally, the removal of themultistorefield and the introduction ofcommitSnapshotterandstorageSnapshotterfields should be carefully documented to inform users of the changes.store/snapshots/helpers_test.go (9)
18-22: The addition of new imports suggests that the test file now depends on additional modules, which is expected given the changes described in the pull request summary. Ensure that these new dependencies are properly managed and versioned.
109-111: The
mockCommitSnapshotterstruct has been introduced to mock the behavior of a commit snapshotter. This is a common testing pattern and seems to be correctly implemented here.113-117: The
Restoremethod checks for an unknown format and returns an error if the format is not recognized. This is a good practice for error handling and validation of input data.140-146: The
Snapshotmethod correctly iterates over items and writes them using thesnapshottypes.WriteExtensionPayloadfunction. This is consistent with the expected behavior of a snapshotter.152-158: The
SnapshotFormatandSupportedFormatsmethods return the current format, which is expected behavior for a snapshotter implementation.160-164: The
mockStorageSnapshotterstruct is a stub with a no-opRestoremethod. This is acceptable for a mock implementation, but ensure that the actual implementation ofStorageSnapshotterinterface handles the restoration logic as required.166-186: The
mockErrorCommitSnapshotterstruct is designed to simulate errors during snapshot operations. This is useful for testing error handling in the snapshot manager.149-198: > Note: This review was outside of the patch, so it was mapped to the patch with the greatest overlap. Original lines [188-208]
The
setupBusyManagerfunction sets up a manager that is busy creating a snapshot. This is a good use of goroutines and channels to simulate concurrent operations in tests.
- 215-242: The
hungCommitSnapshottertype is used to simulate a snapshotter that is "hung" or taking an indefinite amount of time to complete an operation. This is achieved by blocking on a channel read, which can be unblocked by calling theClosemethod. This is a good pattern for testing timeout and cancellation behavior.Overall, the changes in the test file align with the new snapshot manager feature and provide mocks and utilities for testing the new functionality. Ensure that all new code paths are covered by tests and that the tests are passing.
store/commitment/iavl/tree.go (8)
20-20: The
IavlTreestruct now implements thesnapshottypes.CommitSnapshotterinterface, which is a significant change to ensure that theIavlTreecan handle snapshot operations. This change should be thoroughly tested to ensure that the implementation is correct and does not introduce any regressions.26-27: The addition of the
storeKeyfield to theIavlTreestruct is a notable change. It is important to ensure that this field is properly set and used throughout the codebase, especially in the context of snapshot operations where the store name is critical.31-36: The
NewIavlTreefunction now accepts astoreKeyparameter. This change will require updates to all places in the code whereNewIavlTreeis called. It is important to verify that thestoreKeyis correctly passed and handled in all instances to avoid any potential issues with store identification.end hunk 0
<<<<<<< start hunk 1
96-151: The
Snapshotmethod has been implemented to create a snapshot of theIavlTreeat a specific version and write it to a protobuf writer. This method should be carefully reviewed to ensure that the snapshot creation process is correct and efficient. The use ofprotoWriter.WriteMsgto write the snapshot items must be correct to ensure the snapshot can be restored properly. Additionally, the error handling and the use ofdeferto close the exporter are good practices to prevent resource leaks.154-228: The
Restoremethod has been implemented to restore theIavlTreefrom a snapshot. This method reads from a protobuf reader and writes key-value pairs to a channel for storage restoration. It is crucial to ensure that the restoration process is accurate and that the error handling is robust. The use of a loop with aswitchstatement to handle different snapshot item types is a common pattern, but it must be ensured that all cases are handled correctly and that the loop exits as expected. Theimporter.Commitandimporter.Closecalls are critical and should be reviewed to ensure that they are correctly implemented to avoid any data corruption or resource leaks.170-171: The
storeKeyis being set from the snapshot item, which implies that thestoreKeyis expected to be part of the snapshot. This should be verified to ensure that the snapshot format includes the store name and that it is correctly set during the snapshot creation process.202-206: The code is sending a
store.KVPairto thechStoragechannel, which is part of the restoration process. It is important to ensure that the receiving end of this channel is correctly processing these key-value pairs and that there is proper synchronization to handle concurrent sends if theRestoremethod is called concurrently.225-227: After the restoration process, the code attempts to load the version of the tree that was just restored. It is important to verify that this operation is successful and that the tree state is consistent after a restore operation.
end hunk 1
store/snapshots/types/snapshotter.go
Outdated
There was a problem hiding this comment.
do we need a types sub-package? Why not just have it all under snapshots. Generally, I've been trying to avoid the use of a types package.
There was a problem hiding this comment.
may I ask you about the reason? I think types package is useful to avoid the import cycle
There was a problem hiding this comment.
"types" is typically an anti-pattern in Golang. It ends up becoming a dumpster of random unrelated stuff (e.g. types in the SDK).
I have all interfaces and basic types in the root store package.
There was a problem hiding this comment.
Review Status
Actionable comments generated: 5
Configuration used: CodeRabbit UI
Files selected for processing (3)
- store/commitment/iavl/tree.go (2 hunks)
- store/snapshots/manager.go (6 hunks)
- store/storage/sqlite/db.go (4 hunks)
Additional comments: 11
store/snapshots/manager.go (4)
35-46: The
Managerstruct has been refactored to includecommitSnapshotterandstorageSnapshotterfields, and themultistorefield has been removed. This change aligns with the new snapshot manager feature and should be checked for compatibility with existing code that uses theManagerstruct.77-90: The
NewManagerfunction has been updated to acceptcommitSnapshotterandstorageSnapshotterparameters. This change should be verified across the codebase to ensure that all instantiations ofManagerare updated accordingly.203-206: The
createSnapshotfunction now uses thecommitSnapshotterto create a snapshot. This change should be tested to ensure that the snapshot creation process works correctly with the newcommitSnapshotterimplementation.362-381: > Note: This review was outside of the patch, so it was mapped to the patch with the greatest overlap. Original lines [362-415]
The
doRestoreSnapshotfunction has been updated to use thestorageSnapshotterfor restoring the storage state. It's important to ensure that the error handling and channel synchronization logic are correctly implemented to prevent data races or deadlocks.The overall changes to the
Managerstruct and related functions appear to be in line with the introduction of the new snapshot manager feature. It is crucial to perform thorough testing, especially for the new snapshot and restore functionalities, to ensure that they work as expected and are compatible with the existing snapshot format. Additionally, the removal of themultistorefield and the introduction ofcommitSnapshotterandstorageSnapshotterfields should be carefully checked for compatibility with existing code.store/commitment/iavl/tree.go (7)
3-17: The addition of new imports should be verified to ensure they are used within the file and there are no unused imports, which can lead to unnecessary package dependencies.
21-21: The implementation of the
snapshottypes.CommitSnapshotterinterface should be thoroughly tested to ensure that theSnapshotandRestoremethods work correctly with the existing snapshot format.27-28: The addition of the
storeKeyfield to theIavlTreestruct is a significant change. Ensure that all instances ofIavlTreecreation are updated to include this new parameter.94-95: The modification of the
Prunemethod to return an error is a change in the method signature that could affect error handling logic in the calling code. Ensure that all calls toPrunehandle the returned error appropriately.97-153: The new
Snapshotmethod should be reviewed for correctness and performance. Ensure that the snapshot creation process is efficient and does not introduce performance bottlenecks, especially for large datasets.156-231: The new
Restoremethod should be reviewed for correctness, especially the error handling and resource management (e.g., ensuring thatimporteris closed properly withdefer).233-235: If the
Closemethod is intended to release resources, verify that it does so correctly. If there are no resources to release, consider whether the method is necessary.---existing comments on 103-103---
alexanderbez's suggestion to uselatestVersion := t.GetLatestVersion()is already implemented in the current code, so no action is needed here.
|
I will add the abstraction layer to both commitment and storage. It is for multi stores in the commitment and avoiding duplication in the storage. |
There was a problem hiding this comment.
Review Status
Actionable comments generated: 1
Configuration used: CodeRabbit UI
Files selected for processing (1)
- store/snapshots/manager.go (6 hunks)
Additional comments: 5
store/snapshots/manager.go (5)
35-46: The refactoring of the
Managerstruct to includecommitSnapshotterandstorageSnapshotterwhile removingmultistoreis a significant change. Ensure that all references tomultistorein the codebase have been updated to use the new fields.166-174: > Note: This review was outside of the patch, so it was mapped to the patch with the greatest overlap. Original lines [166-210]
The
Createmethod has been updated to usecommitSnapshotterfor snapshot creation. Ensure that thecommitSnapshotteris properly initialized before this method is called, and that error handling is robust, especially in the goroutine that generates snapshot chunks.
204-210: The
createSnapshotmethod now usescommitSnapshotter.Snapshot. It's important to ensure that thestreamWriteris properly closed in all cases to avoid resource leaks. The current implementation seems to handle this correctly withdeferandCloseWithError.366-385: > Note: This review was outside of the patch, so it was mapped to the patch with the greatest overlap. Original lines [366-419]
The
doRestoreSnapshotmethod has been refactored to use bothcommitSnapshotterandstorageSnapshotter. It's crucial to ensure that the error channels and goroutines are managed correctly to prevent deadlocks or missed errors. The current implementation appears to correctly handle channel operations and goroutine synchronization.
- 409-416: The error handling for the
storageSnapshotter.Restoremethod within the goroutine should be verified to ensure that any errors are properly propagated and handled. The current implementation seems to correctly use a channel to communicate errors back to the main goroutine.
There was a problem hiding this comment.
Review Status
Actionable comments generated: 0
Configuration used: CodeRabbit UI
Files selected for processing (1)
- store/pruning/manager_test.go (3 hunks)
Additional comments: 3
store/pruning/manager_test.go (3)
26-37: The
SetupTestfunction has been updated to use a new logger for theiavl.IavlTreeand theManager. This change is likely for better test logging and diagnostics. Ensure that the logger used here is appropriate for the test environment and that it does not interfere with the test results or output.47-62: The
TestPruningfunction now includes a loop to dynamically generate key-value pairs for each version, which is a good practice for creating more robust tests. However, ensure that thekvCountvariable is set to a representative number of key-value pairs that would be expected in a real-world scenario to ensure the test is meaningful.70-88: The assertions have been updated to reflect the dynamic generation of key-value pairs. The test now checks for the existence of a key at a specific version and the absence of a key at another version, which is a good practice for testing pruning functionality. Ensure that the versions and keys checked are consistent with the expected behavior of the pruning logic.
There was a problem hiding this comment.
Nice work and thank you @cool-develope, just some suggestions please take a look.
| return fmt.Errorf("failed to get immutable tree for version %d: %w", version, err) | ||
| } | ||
|
|
||
| exporter, err := tree.Export() |
There was a problem hiding this comment.
Could we please call this itr or iter instead of exporter? exporter makes it seem like its purpose is to write out content, but really the purpose is to iterate over the exported content.
store/pruning/manager_test.go
Outdated
| for i := uint64(0); i < latestVersion; i++ { | ||
| version := i + 1 | ||
| cs := store.NewChangeset() | ||
| cs.Add([]byte("key"), []byte(fmt.Sprintf("value%d", version))) | ||
| for j := 0; j < kvCount; j++ { | ||
| cs.Add([]byte(fmt.Sprintf("key-%d", j)), []byte(fmt.Sprintf("value-%d-%d", version, j))) |
There was a problem hiding this comment.
Won't this cause a conflict of keys written given than they'll go kvCount times but be repeated latestVersion times?
There was a problem hiding this comment.
The combination of key and version is unique in the storage, it is enough to test pruning.
| func (m *hungCommitSnapshotter) Restore( | ||
| height uint64, format uint32, protoReader protoio.Reader, chStorage chan<- *store.KVPair, |
There was a problem hiding this comment.
Stylistic nit for unused functions parameters to even indicate to outside callers that it is not to be used
func (m *hungCommitSnapshotter) Restore(_ uint64, _ uint32, _ protoio.Reader, _ chan<- *store.KVPair) (...) {
panic("not implemented")
}
store/commitment/iavl/tree.go
Outdated
| } | ||
|
|
||
| // NewIavlTree creates a new IavlTree instance. | ||
| func NewIavlTree(db dbm.DB, logger log.Logger, cfg *Config) *IavlTree { | ||
| func NewIavlTree(db dbm.DB, logger log.Logger, storeKey string, cfg *Config) *IavlTree { |
There was a problem hiding this comment.
i thought there was discussion around abstracting storekey away? does this change the direction of that conversation?
There was a problem hiding this comment.
I think the expectation is that dbm.DB is a PrefixDb (for iavl v0 & v1) so the store key prefix is already baked into that. I guess storekey is required for the snapshot API? (still reading this PR).
There was a problem hiding this comment.
db dbm.DB is not required or used in iavl v2 (since it's an interface to an underlying generic KV store). Maybe we should fold this into Config if we want to support multiple iavl versions in store v2?
| Restore(height uint64, format uint32, protoReader protoio.Reader) (SnapshotItem, error) | ||
| // CommitSnapshotter defines an API for creating and restoring snapshots of the | ||
| // commitment state. | ||
| type CommitSnapshotter interface { |
There was a problem hiding this comment.
why is there a commitment specific interface here? this seems like an implementation detail that is leaking into an interface
|
It looks like IAVL |
store/storage/sqlite/db.go
Outdated
| @@ -19,6 +20,9 @@ const ( | |||
| reservedStoreKey = "_RESERVED_" | |||
| keyLatestHeight = "latest_height" | |||
|
|
|||
| // TODO: it is a random number, need to be tuned | |||
| defaultBatchBufferSize = 100000 | |||
There was a problem hiding this comment.
For the best snapshot import performance we should consider tuning the max WAL size to something big, on the order of single digit gigabytes. see: https://www.sqlite.org/pragma.html#pragma_wal_autocheckpoint
There was a problem hiding this comment.
also synchronous=off will boost throughput: https://www.sqlite.org/pragma.html#pragma_synchronous
store/snapshots/types/snapshotter.go
Outdated
There was a problem hiding this comment.
"types" is typically an anti-pattern in Golang. It ends up becoming a dumpster of random unrelated stuff (e.g. types in the SDK).
I have all interfaces and basic types in the root store package.
store/commitment/store.go
Outdated
| } | ||
|
|
||
| for storeKey, tree := range c.multiTrees { | ||
| exporter, err := tree.Export(version) |
There was a problem hiding this comment.
This looks like an expensive operation being ran in a loop multiple times. Couldn't we extract the exporter right above once? We could invert the operations that iterate over storeKeys and exporter.Next() to iterate N times over exporter.Next() but storeKeys (M) times
There was a problem hiding this comment.
we have several trees
cool-develope
force-pushed
the
store/snapshot
branch
from
64dc3c9
to
0598995
Compare
There was a problem hiding this comment.
Review Status
Actionable comments generated: 2
Configuration used: CodeRabbit UI
Files selected for processing (24)
- store/changeset.go (1 hunks)
- store/commitment/iavl/exporter.go (1 hunks)
- store/commitment/iavl/importer.go (1 hunks)
- store/commitment/iavl/tree.go (1 hunks)
- store/commitment/iavl/tree_test.go (1 hunks)
- store/commitment/store.go (2 hunks)
- store/commitment/store_test_suite.go (1 hunks)
- store/commitment/tree.go (2 hunks)
- store/kv/branch/store_test.go (3 hunks)
- store/pruning/manager_test.go (3 hunks)
- store/root/store_test.go (2 hunks)
- store/snapshots/chunk.go (2 hunks)
- store/snapshots/helpers_test.go (9 hunks)
- store/snapshots/manager.go (7 hunks)
- store/snapshots/snapshotter.go (1 hunks)
- store/storage/database.go (1 hunks)
- store/storage/pebbledb/db.go (4 hunks)
- store/storage/pebbledb/db_test.go (1 hunks)
- store/storage/rocksdb/db.go (4 hunks)
- store/storage/rocksdb/db_test.go (2 hunks)
- store/storage/sqlite/db.go (4 hunks)
- store/storage/sqlite/db_test.go (4 hunks)
- store/storage/storage_bench_test.go (2 hunks)
- store/storage/store.go (1 hunks)
Additional comments: 67
store/changeset.go (1)
- 6-10: The addition of the
StoreKeyfield to theKVPairstruct is consistent with the PR objectives and the provided summaries. This field is intended to be optional for snapshot restore functionality. Ensure that the rest of the codebase that usesKVPairis aware of this new field and handles it appropriately.store/commitment/iavl/exporter.go (1)
- 12-40: The implementation of the
Exporterstruct and its methodsNextandClosealigns with the PR objectives to introduce a snapshot manager and the changes described in the summary. The error handling in theNextmethod correctly translates theiavl.ErrorExportDonetocommitment.ErrorExportDone, and theClosemethod properly delegates the call to the underlyingiavl.Exporter.store/commitment/iavl/importer.go (1)
- 9-34: The
Importerstruct and its methodsAdd,Commit, andCloseare correctly implemented as wrappers around theiavl.Importermethods. The conversion fromsnapshotstypes.SnapshotIAVLItemtoiavl.ExportNodein theAddmethod is done properly, ensuring that the fields match between the two types.store/commitment/iavl/tree.go (5)
86-86: The concern raised by odeke-em about returning a nil exporter if there is a non-nil error seems to have been addressed, as the code correctly returns
nilalong with the error.98-98: The concern raised by odeke-em and alexanderbez about returning a nil importer for a non-nil error has been addressed, as the code correctly returns
nilalong with the error.91-93: The
Exportertype is correctly instantiated and returned, aligning with the naming convention explained by kocubinski in the previous discussion.103-105: The
Importertype is correctly instantiated and returned, and there are no outstanding concerns from previous reviews regarding this part of the code.77-110: The implementation of the
ExportandImportmethods in theIavlTreestruct appears to be correct and in line with the PR objectives and previous review comments.store/commitment/iavl/tree_test.go (3)
14-27: The addition of
TestCommitterSuiteis a good practice for ensuring that the new snapshot manager functionality is thoroughly tested.30-34: The summary indicates that
generateTreewas renamed, but the code shows it still exists with a modified implementation. This should be clarified in the summary.36-36: The
TestIavlTreefunction provides a comprehensive test for theIavlTreeoperations, which is crucial for ensuring the integrity of the snapshot manager's functionality.store/commitment/store.go (4)
3-21: The addition of new imports and the implementation of the
CommitSnapshotterinterface by theCommitStoretype align with the PR objectives to introduce a snapshot manager.138-195: The
Snapshotmethod correctly checks for version validity and ensures the snapshot version is not greater than the latest version. However, the TODO comment about checking the parallelism of the loop suggests that there may be plans to parallelize this operation. It's important to ensure thread safety and data consistency if parallel execution is implemented in the future.198-276: The
Restoremethod appears to be well-implemented with proper error handling and validation checks. The use of a channel (chStorage) for storage operations is a good practice for concurrent operations, which aligns with the PR's goal of introducing a snapshot manager that works with the existing snapshot format.278-280: The
Closemethod properly attempts to close all trees and collects any errors that occur usingerrors.Join. This is a good practice for resource management and error reporting.store/commitment/store_test_suite.go (2)
- 29-120: The test suite effectively covers the snapshot and restore functionality of the
CommitStore. It tests the entire process of creating a snapshot, writing an extension metadata, and restoring from the snapshot, ensuring that the restored data matches the original data. The use of channels and goroutines for streaming data is appropriate and the synchronization withsync.WaitGroupis correctly implemented to wait for the goroutine to finish processing. The test also verifies that the hash of the restored tree matches the original, which is crucial for ensuring data integrity.The test suite seems to be comprehensive and well-structured, with proper error handling and assertions. However, it is important to ensure that the test suite is integrated with the rest of the application's tests and that it is executed as part of the continuous integration process to catch any regressions or issues early on.
- 29-120: While the unit tests provided in this suite are comprehensive, it is important to remember that end-to-end tests are not included and should be added during the integration with the base application to ensure that the snapshot manager works correctly within the full context of the application.
store/commitment/tree.go (4)
12-13: The introduction of
ErrorExportDoneis clear and follows Go conventions for error declaration.22-44: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [15-29]
The extension of the
Treeinterface withExportandImportmethods aligns with the PR's objective to introduce snapshot management features.
31-36: The
Exporterinterface is correctly defined with theNextmethod, which is expected to iterate over items to be exported.38-44: The
Importerinterface is correctly defined withAddandCommitmethods to facilitate the import of snapshot items.store/kv/branch/store_test.go (3)
8-13: The addition of the
"cosmossdk.io/store/v2/storage"import aligns with the changes made in theSetupTestfunction to use the newstoragepackage.28-31: The
SetupTestfunction has been updated to use the newstoragepackage, which is consistent with the PR objectives to introduce a new snapshot manager and related functionalities.41-47: The
ApplyChangesetmethod is now called on thessvariable, which is an instance ofStorageStore, and thekvStoreis initialized withss. This reflects the changes made to use the newstoragepackage.store/pruning/manager_test.go (3)
11-15:
The addition of new imports for"cosmossdk.io/store/v2/storage"and"cosmossdk.io/store/v2/storage/sqlite"is consistent with the PR objectives to support the new snapshot manager feature.38-43:
The renaming ofsstosqliteDBand the creation of a newstorage.NewStorageStoreinstance are in line with the PR's goal to implement a snapshotter API for storage stores.63-66:
The comment change from "write 10 batches" to "write batches" is a minor clarification and does not impact the test's logic.store/root/store_test.go (1)
- 37-37: Verify that
defaultStoreKeyis defined and used consistently across the codebase.
Verification result:
The
defaultStoreKeyis properly defined instore/pruning/manager_test.goandstore/root/store.goand is used consistently across the codebase in various test files and the main store implementation.store/snapshots/chunk.go (3)
- 6-9: The import path
snapshottypeshas been correctly renamed tosnapshotstypes. Ensure that all references to the old import path are updated throughout the entire codebase.
Verification result:
The import path change from
snapshottypestosnapshotstypeshas been successfully reflected in the codebase. The search for the old import path returned no results, indicating that it has been fully replaced. The new import path is correctly used in multiple files, as verified by the search results.
170-182: The function
ValidRestoreHeightcorrectly uses the renamed importsnapshotstypes. This change is consistent with the PR objectives and the summary.169-183: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [1-183]
No further issues found in the provided file. Ensure that the renaming of the import path is reflected across all files that may use it.
Verification result:
The renaming of the import path from
snapshottypestosnapshotstypeshas been successfully reflected in multiple files across the codebase, as indicated by the search results. No instances of the old import path were found, suggesting that the necessary updates have been made.store/snapshots/helpers_test.go (3)
238-242: The
Restoremethod inhungCommitSnapshotteris not implemented and panics. If this is intentional to indicate that the method should not be used, it's fine. However, if this method is expected to be implemented in the future, it would be helpful to add a TODO comment or create an issue to track this.270-290: The
SnapshotExtensionandRestoreExtensionmethods inextSnapshotterhave been updated to use thesnapshotspackage. Ensure that these changes are consistent with the intended use of thesnapshotspackage and that all necessary functionality is present.17-25: > Note: This review was outside the patches, and no patch overlapping with it was found. Original lines [291-308]
The
GetTempDirfunction has been updated to useos.MkdirTempinstead oftesting.T.TempDir. This change is not mentioned in the summary. Verify that this change is intentional and does not affect the test behavior.store/snapshots/manager.go (8)
31-45: The restructuring of the
Managerstruct with the removal ofmultistoreand addition ofcommitSnapshotterandstorageSnapshotteraligns with the PR objectives to introduce a snapshot manager that works with the existing snapshot format. The changes to theextensionsmap andloggerfield types are also noted and appear to be consistent with the rest of the codebase.77-90: The
NewManagerfunction has been correctly updated to reflect the changes in theManagerstruct, ensuring that the new fieldscommitSnapshotterandstorageSnapshotterare initialized. The handling ofnilextensions is also appropriate, providing a default empty map if none are provided.92-98: The
RegisterExtensionsmethod has been updated to use the newExtensionSnapshottertype, which is consistent with the changes in theManagerstruct. The error handling for duplicate snapshotter names and unsupported snapshot formats is also appropriate.166-174: The
Createmethod has been updated to use thecommitSnapshotterfor snapshot creation, which is in line with the PR objectives to implement a snapshotter API for both commitment and storage stores. The error handling and defer statement to end the operation are correctly implemented.204-210: The
createSnapshotmethod correctly delegates the snapshot creation to thecommitSnapshotter, and the error handling withCloseWithErroris appropriate to ensure that any errors during the snapshot creation are propagated correctly.366-385: The
doRestoreSnapshotmethod has been updated to use thestorageSnapshotterfor restoring the storage state, which aligns with the PR objectives. The channelchStorageis used to pass the key-value pairs to thestorageSnapshotter, and the error handling is consistent with the rest of the codebase.409-418: The
doRestoreSnapshotmethod correctly handles the restoration of extensions and waits for thestorageSnapshotterto complete. The error handling for the storage snapshotter is appropriate, ensuring that any errors are captured and returned.517-523: The
IsFormatSupportedfunction correctly checks if the snapshotter supports the given format, which is a necessary utility function for the snapshot manager to verify compatibility with different snapshot formats.store/snapshots/snapshotter.go (2)
10-24: The interfaces
CommitSnapshotterandStorageSnapshotterhave been correctly defined to separate the concerns of commitment state and storage state snapshotting. This aligns with the PR objectives to introduce a snapshot manager that works with the existing snapshot format without breaking it.26-26: The
ExtensionSnapshotterinterface remains unchanged and is not part of the PR's objectives to modify snapshotting functionality.store/storage/database.go (2)
9-23: The
Databaseinterface is well-defined and aligns with the PR objectives to provide an abstraction layer for storage operations, including versioning and iteration capabilities. It is important to ensure that all implementations of this interface are consistent with the methods defined here.9-23: While the
Databaseinterface is well-defined, it is important to note that end-to-end tests for the new snapshot manager feature, including the implementations of this interface, are not included in this PR. It should be ensured that these tests are planned and tracked for future implementation to verify the integration with the base application.store/storage/pebbledb/db.go (3)
11-15: The addition of the
cosmossdk.io/store/v2/storageimport is consistent with the changes made to theDatabasestruct and the new snapshot manager functionality.29-32: The change from
store.VersionedDatabasetostorage.Databasein the type assertion for theDatabasestruct aligns with the new snapshot manager's API and storage abstraction layer.96-102: The new
NewBatchmethod added to theDatabasestruct is consistent with the snapshotter API implementation for storage stores. Ensure that theNewBatchmethod is correctly integrated with the rest of the system and that its usage is covered by tests, especially since the PR notes that end-to-end tests are not included.
Verification result:
The
NewBatchmethod has been successfully integrated into theDatabasestruct within thestore/storage/pebbledb/db.gofile, aligning with the snapshotter API for storage stores. The method's usage is confirmed in various parts of the system, includingstore/storage/store.go,store/storage/rocksdb/db.go,store/storage/sqlite/db.go, andstore/storage/pebbledb/db.go. However, it is important to ensure that this integration is accompanied by adequate unit tests, and the absence of end-to-end tests as noted in the PR should be addressed in subsequent updates.store/storage/pebbledb/db_test.go (1)
- 19-25: The changes in the
TestStorageTestSuitefunction reflect the introduction of the new storage abstraction layer, as theNewDBfunction now wraps thepebbledbdatabase withstorage.NewStorageStore. This aligns with the PR's objective to implement a snapshotter API for storage stores. Ensure that theSetSync(false)call and theEmptyBatchSizeconfiguration are appropriate for the test environment and consistent with the intended usage patterns.store/storage/rocksdb/db.go (4)
27-27: Ensure that the
Databasestruct implements all methods required by thestorage.Databaseinterface to maintain compatibility.94-95: Confirm that the
NewBatchfunction is defined and correctly creates a batch with the given version.91-100: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [12-98]
Verify that the removal of the
ApplyChangesetmethod is intentional and assess the impact on other components that might use this method.
- 91-100: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [1-149]
The review is limited to the
store/storage/rocksdb/db.gofile and does not cover changes to theIavlTreestructure or theManagerstruct mentioned in the summary.store/storage/rocksdb/db_test.go (3)
21-28: The update to the
NewDBmethod in theTestStorageTestSuitefunction correctly reflects the new signature and expected return type.34-45: The changes in the
TestDatabase_ReverseIteratorfunction, including the use ofNewBatchandbatch.Write(), are consistent with the new storage snapshotter API implementation.47-47: Ensure that the
ReverseIteratormethod is correctly implemented and tested, as it is crucial for the snapshot restoration process.store/storage/sqlite/db.go (3)
11-15: The addition of the
"cosmossdk.io/store/v2/storage"import aligns with the implementation of thestorage.Databaseinterface by theDatabasestruct.44-44: The
Databasestruct correctly asserts that it implements thestorage.Databaseinterface, which is consistent with the PR's objective to introduce a snapshot manager.95-97: The addition of the
NewBatchmethod to theDatabasestruct is a new feature that aligns with the snapshot manager's functionality.store/storage/sqlite/db_test.go (1)
- 19-26: : Please confirm that setting
EmptyBatchSizeto 0 inTestStorageTestSuiteis intentional and aligns with the expected behavior for the storage tests.store/storage/storage_bench_test.go (5)
4-4: The change in package declaration to
storage_testensures that the tests are black-box tests, which is a good practice in Go.27-40: The
backendsmap now wraps the database instances withstorage.NewStorageStore, aligning with the new snapshot manager feature.1-7: > Note: This review was outside the patches, and no patch overlapping with it was found. Original lines [42-42]
The benchmarks correctly use the updated
backendsmap to test the new storage abstraction.
22-23: The introduction of the
storeKey1constant is likely for simulating a store key in the benchmarks, which aligns with the new snapshot manager feature.41-41: The use of
rngfor generating random keys and values in benchmarks is appropriate and common practice.store/storage/store.go (1)
- 20-130: The implementation of
StorageStorealigns with the PR objectives to provide an abstraction layer for storage operations and implement the snapshotter API. The methods are consistent with the expected functionality of a storage abstraction in a snapshot manager system.
| func (i *Importer) Close() error { | ||
| i.importer.Close() | ||
|
|
||
| return nil |
There was a problem hiding this comment.
The Close method does not handle the error that might be returned by i.importer.Close(). It is a best practice to handle errors when closing resources to ensure that any issues are properly logged or managed.
func (i *Importer) Close() error {
- i.importer.Close()
+ if err := i.importer.Close(); err != nil {
+ return err
+ }
return nil
}Committable suggestion
❗ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation.
| func (i *Importer) Close() error { | |
| i.importer.Close() | |
| return nil | |
| func (i *Importer) Close() error { | |
| if err := i.importer.Close(); err != nil { | |
| return err | |
| } | |
| return nil | |
| } |
| // TODO: it is a random number, need to be tuned | ||
| defaultBatchBufferSize = 100000 |
There was a problem hiding this comment.
Consider determining an optimal value for defaultBatchBufferSize or providing a way to configure it dynamically based on the workload or environment.
| i.importer.Close() | ||
|
|
||
| return nil |
There was a problem hiding this comment.
| i.importer.Close() | |
| return nil | |
| return i.importer.Close() |
There was a problem hiding this comment.
iavl importer returns nothing
| // TODO: check the parallelism of this loop | ||
| if err := func() error { |
There was a problem hiding this comment.
Why is this in a function closure?
There was a problem hiding this comment.
it is for defer expoerter.Close()
| } | ||
| }() | ||
|
|
||
| nextItem, err = m.commitSnapshotter.Restore(snapshot.Height, snapshot.Format, streamReader, chStorage) |
There was a problem hiding this comment.
Question: why not restore SS and SC both async?
e.g.
eg := new(errgroup.Group)
eg.Go(func() error { /* restore SS */ })
eg.Go(func() error { /* restore SC */ })
if err := eg.Wait(); err != nil {
return fmt.Errorf("failed to restore: %w", err)
}
return nilThere was a problem hiding this comment.
SC restore is using the same routine of doRestoreSnapshot, and it returns nextItem which will be used in the extension restoring.
store/snapshots/manager.go
Outdated
|
|
||
| snapshotMaxItemSize = int(64e6) // SDK has no key/value size limit, so we set an arbitrary limit | ||
| ) | ||
|
|
||
| var ErrOptsZeroSnapshotInterval = errors.New("snaphot-interval must not be 0") | ||
|
|
||
| // NewManager creates a new manager. | ||
| func NewManager(store *Store, opts types.SnapshotOptions, multistore types.Snapshotter, extensions map[string]types.ExtensionSnapshotter, logger log.Logger) *Manager { | ||
| func NewManager(store *Store, opts types.SnapshotOptions, commitSnapshotter CommitSnapshotter, storageSnapshotter StorageSnapshotter, extensions map[string]ExtensionSnapshotter, logger log.Logger) *Manager { |
There was a problem hiding this comment.
Shouldn't we move types/options.go to the root as well? What is the point of types package?
There was a problem hiding this comment.
it is a original snapshot structure
There was a problem hiding this comment.
Review Status
Actionable comments generated: 1
Configuration used: CodeRabbit UI
Files selected for processing (4)
- store/snapshots/manager.go (7 hunks)
- store/snapshots/manager_test.go (8 hunks)
- store/snapshots/options.go (1 hunks)
- store/storage/database.go (1 hunks)
Additional comments: 18
store/snapshots/manager.go (8)
31-43: The changes to the
Managerstruct, including the addition ofcommitSnapshotter,storageSnapshotter, andloggerfields, align with the PR objectives to introduce a snapshot manager for the Cosmos SDK store system. These fields are essential for managing the snapshotting process for both the commitment and storage store as described.65-90: The updates to constants
chunkBufferSize,chunkIDBufferSize, anddefaultStorageChannelBufferSizeare consistent with the need to manage buffer sizes for snapshot operations. TheNewManagerfunction now correctly takescommitSnapshotterandstorageSnapshotteras parameters, which is in line with the PR's goal to implement a snapshotter API.166-174: The
Createmethod's changes, including the removal of the call tomultistore.PruneSnapshotHeightand the use ofcommitSnapshotter, align with the PR's objectives to update the snapshot management logic.204-210: The
createSnapshotmethod's changes, specifically the use ofcommitSnapshotterto create a snapshot, are in line with the PR's objectives to implement a snapshotter API for the commitment store.366-385: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [366-419]
The
doRestoreSnapshotmethod's changes, including the introduction of a channelchStorageto pass KV pairs to the storage snapshotter and the error handling logic, are consistent with the PR's objectives to implement a snapshotter API for the storage store.
517-523: The
IsFormatSupportedfunction's changes, including the acceptance ofExtensionSnapshotterinstead oftypes.ExtensionSnapshotter, align with the PR's objectives to update the snapshot interfaces.68-70: The previous comment regarding the introduction of
defaultStorageChannelBufferSizeas a global variable remains valid. The buffer size should be appropriate for the expected workload and not lead to excessive memory consumption.382-382: The conversation between
alexanderbezandcool-developeregarding the asynchronous restoration of SS and SC has been considered. The current implementation reflects a design decision to use the same routine for SC restore asdoRestoreSnapshot, which returnsnextItemfor extension restoring.store/snapshots/manager_test.go (7)
16-16: The summary mentions the use of
types.NewSnapshotOptionsbefore the changes, but the provided code only showssnapshots.NewSnapshotOptions. Please verify if this is an inconsistency or if the summary is incorrect.74-76: The summary does not mention the addition of
manager.RegisterExtensions(extSnapshotter)calls in the test cases. This seems to be an important part of the test setup that should be included in the summary.24-25: The summary does not mention the use of
setupBusyManager(t)to simulate a busy manager state in the tests. This seems to be an important part of the test logic that should be included in the summary.72-72: The summary does not mention the use of
newExtSnapshotter(10)andnewExtSnapshotter(0)in the test cases. This seems to be an important part of the test setup that should be included in the summary.95-95: The summary does not mention the use of
types.Metadata{ChunkHashes: checksums(chunks)}in the test cases. This seems to be an important part of the test assertions that should be included in the summary.181-186: The summary does not mention the use of
manager.RestoreChunk(chunk)in the test cases. This seems to be an important part of the test logic that should be included in the summary.239-239: The summary does not mention the use of
mockErrorCommitSnapshotterin the test cases. This seems to be an important part of the test logic that should be included in the summary.store/snapshots/options.go (1)
- 1-4: > Note: This review was outside the patches, so it was mapped to the patch with the greatest overlap. Original lines [3-14]
The
SnapshotOptionsstruct andNewSnapshotOptionsfunction are correctly implemented and do not exhibit any issues.store/storage/database.go (2)
9-25: The
Databaseinterface provides a clear abstraction for storage operations, which aligns with the PR objectives to introduce a storage abstraction layer. The methods cover batch creation, key existence checks, value retrieval, version management, iteration, pruning, and closing operations, which are essential for the snapshot manager's functionality.9-25: Verify if the
Databaseinterface introduces duplication with the existingVersionedDatabasein the rootstorepackage, as mentioned by alexanderbez. If it does, consider refactoring to extend or useVersionedDatabaseto avoid code duplication.
Verification result:
The
Databaseinterface does not appear to introduce duplication with theVersionedDatabaseinterface. The search results indicate thatVersionedDatabaseis used in various contexts throughout the codebase, and theDatabaseinterface is a new addition that may serve a different purpose or provide a more specific abstraction for the storage system.
| package snapshots | ||
|
|
There was a problem hiding this comment.
💡 NOTE
Codebase verification is a beta feature.
The package name change from types to snapshots in store/snapshots/options.go has been confirmed. However, the search for the old package name types indicates that numerous files across various modules still reference the old package name. These references need to be updated to reflect the new package name. The search for the new package name snapshots confirms that several files within the store/snapshots directory and one file in the api/cosmos/store/snapshots/v1 directory have been updated accordingly.
- Update all references from
package typestopackage snapshotsacross the codebase as indicated by the search results. - Verify that the changes do not conflict with other modules that may also use a
typespackage.
Original review comment
The package name has been changed to snapshots. Ensure that all references to this package across the codebase are updated to reflect this change.
| return snapshotstypes.SnapshotItem{}, fmt.Errorf("received IAVL node item before store item") | ||
| } | ||
| node := item.IAVL | ||
| if node.Height > int32(math.MaxInt8) { |
There was a problem hiding this comment.
why do we have this restriction, is it something related to iavl?
There was a problem hiding this comment.
yes, iavl requires int8
There was a problem hiding this comment.
hmm, we should change this in iavlv2, its weird versions can go up to maxint64 but iavl will crash before then. Seems like there is a bit of a misstep in expectations between the two dependencies
There was a problem hiding this comment.
I agree that in v2 of iavl we should remove this restriction. It makes code usage in iavl also awkward: one weekend many months ago before I found out about Matt's rewrite, I tried to tackle concurrency issues and data races in iavl using atomic.Int64 and it was awkward trying to use that in iavl.
There was a problem hiding this comment.
it is a just tree height, it can be up to maxUint8 (255)
| @@ -175,29 +185,6 @@ func (db *Database) Get(storeKey string, targetVersion uint64, key []byte) ([]by | |||
| return nil, nil | |||
| } | |||
|
|
|||
| func (db *Database) ApplyChangeset(version uint64, cs *store.Changeset) error { | |||
There was a problem hiding this comment.
could you explain why this was removed?
There was a problem hiding this comment.
it is moved to store to avoid the duplication ( the reason of abstraction layer in the storage)
There was a problem hiding this comment.
Review Status
Actionable comments generated: 0
Configuration used: CodeRabbit UI
Files selected for processing (1)
- store/root/store_test.go (2 hunks)
Additional comments: 3
store/root/store_test.go (3)
12-17: The addition of the
"cosmossdk.io/store/v2/storage"import aligns with the PR's objective to introduce a new snapshot manager and related functionalities.33-41: The changes in the
SetupTestmethod reflect the PR's objective to implement a storage abstraction layer by creating asqliteDBand then wrapping it withstorage.NewStorageStore.37-39: The
scvariable now usesdefaultStoreKeyas a key in the map, which is a change from the previous version. This should be verified to ensure thatdefaultStoreKeyis correctly defined and used throughout the codebase.
The usage of
defaultStoreKeyis consistent across the codebase, with its definition found instore/root/store.goand usage in various test and store files. No issues found with the introduction ofdefaultStoreKey.
* feat: secp256k1 public key constant time (cosmos#18026) Signed-off-by: bizk <santiago.yanzon1999@gmail.com> * chore: Fixed changelog duplicated items (cosmos#18628) * adr: Un-Ordered Transaction Inclusion (cosmos#18553) Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Aleksandr Bezobchuk <alexanderbez@users.noreply.github.com> * docs: lint ADR-070 (cosmos#18634) * fix(baseapp)!: postHandler should run regardless of result (cosmos#18627) * docs: fix typos in adr-007-specialization-groups.md (cosmos#18635) * chore: alphabetize labels (cosmos#18640) * docs(x/circuit): add note on ante handler (cosmos#18637) Co-authored-by: Aleksandr Bezobchuk <alexanderbez@users.noreply.github.com> * fix: telemetry metric label variable (cosmos#18643) * chore: typos fix (cosmos#18642) * refactor(store/v2): updates from integration (cosmos#18633) * build(deps): Bump actions/setup-go from 4 to 5 (cosmos#18647) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: github-actions <41898282+github-actions[bot]@users.noreply.github.com> * feat(store/v2): snapshot manager (cosmos#18458) * chore(client/v2): fix typos in the README.md (cosmos#18657) * fix(baseapp): protocompat.go gogoproto.Merge does not work with custom types (cosmos#18654) Co-authored-by: unknown unknown <unknown@unknown> * chore: fix several minor typos (cosmos#18660) * chore(tools/confix/cmd): fix typo in view.go (cosmos#18659) * refactor(x/staking): check duplicate addresses in StakeAuthorization's params (cosmos#18655) * feat(accounts): use gogoproto API instead of protov2. (cosmos#18653) Co-authored-by: unknown unknown <unknown@unknown> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> * fix(store/commitment/iavl): honor tree.Remove error firstly (cosmos#18651) * build(deps): Bump actions/stale from 8 to 9 (cosmos#18656) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * chore(docs): fix typos & wording in docs (cosmos#18667) * chore: fix several typos. (cosmos#18666) * feat(telemetry): enable `statsd` and `dogstatsd` telemetry sinks (cosmos#18646) Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Aleksandr Bezobchuk <alexanderbez@users.noreply.github.com> Co-authored-by: marbar3778 <marbar3778@yahoo.com> Co-authored-by: Marko <marko@baricevic.me> * feat(store/v2): add SetInitialVersion in SC (cosmos#18665) * feat(client/keys): support display discreetly for `keys add` (cosmos#18663) Co-authored-by: Julien Robert <julien@rbrt.fr> * ci: add misspell action (cosmos#18671) * chore: typos fix by misspell-fixer (cosmos#18683) Co-authored-by: github-merge-queue <github-merge-queue@users.noreply.github.com> Co-authored-by: Julien Robert <julien@rbrt.fr> * chore: add v0.50.2 changelog to main (cosmos#18682) * build(deps): Bump github.com/jhump/protoreflect from 1.15.3 to 1.15.4 in /tests (cosmos#18678) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> * refactor(bank): remove .String() calls (cosmos#18175) Co-authored-by: Facundo <facundomedica@gmail.com> * ci: use codespell instead of misspell-fixer (cosmos#18686) Co-authored-by: Marko <marbar3778@yahoo.com> * feat(gov): add proposal types and spam votes (cosmos#18532) * feat(accounts): use account number as state prefix for account state (cosmos#18664) Co-authored-by: unknown unknown <unknown@unknown> * chore: typos fixes by cosmos-sdk bot (cosmos#18689) Co-authored-by: github-merge-queue <github-merge-queue@users.noreply.github.com> Co-authored-by: Julien Robert <julien@rbrt.fr> Co-authored-by: marbar3778 <marbar3778@yahoo.com> * feat(client/keys): support display discreetly for keys mnemonic (cosmos#18688) * refactor: remove panic usage in keeper methods (cosmos#18636) * ci: rename pr name in misspell job (cosmos#18693) Co-authored-by: Marko <marko@baricevic.me> * build(deps): Bump github.com/pelletier/go-toml/v2 from 2.1.0 to 2.1.1 in /tools/confix (cosmos#18702) Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: github-actions <41898282+github-actions[bot]@users.noreply.github.com> * feat(client/keys): support display discreetly for keys export (cosmos#18684) * feat(x/gov): better gov genesis validation (cosmos#18707) --------- Signed-off-by: bizk <santiago.yanzon1999@gmail.com> Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: Carlos Santiago Yanzon <27785807+bizk@users.noreply.github.com> Co-authored-by: yihuang <huang@crypto.com> Co-authored-by: coderabbitai[bot] <136622811+coderabbitai[bot]@users.noreply.github.com> Co-authored-by: Aleksandr Bezobchuk <alexanderbez@users.noreply.github.com> Co-authored-by: Facundo Medica <14063057+facundomedica@users.noreply.github.com> Co-authored-by: Akaonetwo <107335783+Akare123@users.noreply.github.com> Co-authored-by: Marko <marbar3778@yahoo.com> Co-authored-by: Julien Robert <julien@rbrt.fr> Co-authored-by: dreamweaverxyz <153101746+dreamweaverxyz@users.noreply.github.com> Co-authored-by: Pioua <136521243+dzizazda@users.noreply.github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: github-actions <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: cool-developer <51834436+cool-develope@users.noreply.github.com> Co-authored-by: leonarddt05 <139609434+leonarddt05@users.noreply.github.com> Co-authored-by: testinginprod <98415576+testinginprod@users.noreply.github.com> Co-authored-by: unknown unknown <unknown@unknown> Co-authored-by: Sukey <35202440+sukey2008@users.noreply.github.com> Co-authored-by: axie <152680487+azukiboy@users.noreply.github.com> Co-authored-by: Luke Ma <867273263@qq.com> Co-authored-by: Emmanuel T Odeke <emmanuel@orijtech.com> Co-authored-by: 0xn4de <109149873+0xn4de@users.noreply.github.com> Co-authored-by: hattizai <150505746+hattizai@users.noreply.github.com> Co-authored-by: Devon Bear <itsdevbear@berachain.com> Co-authored-by: Marko <marko@baricevic.me> Co-authored-by: Halimao <1065621723@qq.com> Co-authored-by: Cosmos SDK <113218068+github-prbot@users.noreply.github.com> Co-authored-by: github-merge-queue <github-merge-queue@users.noreply.github.com> Co-authored-by: Facundo <facundomedica@gmail.com> Co-authored-by: Likhita Polavarapu <78951027+likhita-809@users.noreply.github.com>
|
I'm trying to restore chain's data from the snapshot manager using @yihuang snapshot cli. Here is the reference: #16067 I have posted the error i'm receiving in that PR. Did this recent PR break the snapshot cli? When trying to do a snapshot statesync, I received the following error when starting the chain: |
Description
ref: #18329
NOTE:
Author Checklist
All items are required. Please add a note to the item if the item is not applicable and
please add links to any relevant follow up issues.
I have...
!to the type prefix if API or client breaking changeCHANGELOG.mdmake lintandmake testReviewers Checklist
All items are required. Please add a note if the item is not applicable and please add
your handle next to the items reviewed if you only reviewed selected items.
I have...
!in the type prefix if API or client breaking changeSummary by CodeRabbit
New Features
Bug Fixes
Documentation
Refactor
StorageStoreabstraction.Managerstruct to accommodate new snapshotter interfaces.Style
Tests
Chores