-
Notifications
You must be signed in to change notification settings - Fork 2.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[feat][storage] Add SpanKind support for badger #6376
base: main
Are you sure you want to change the base?
Conversation
I have changed the structure of cache which is leading to these concerns:
Once the correct approach is discussed I will handle some more edge cases and make the e2e tests pass (making |
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #6376 +/- ##
==========================================
- Coverage 96.03% 96.02% -0.01%
==========================================
Files 364 365 +1
Lines 20690 20823 +133
==========================================
+ Hits 19870 19996 +126
- Misses 626 631 +5
- Partials 194 196 +2
Flags with carried forward coverage won't be shown. Click here to find out more. ☔ View full report in Codecov by Sentry. |
@yurishkuro Please review the approach and problems! |
@yurishkuro I have added more changes which reduces the iterations in prefill to 1 but it limits the |
I have an idea for old data without using the migration script! We can store the old data in two other data structures in cache (without kind). But then the only question which rises then: What to return when no span kind is given by user? Operations of new data of all kind or operations of old data (kind marked as unspecified) or an addition of both? |
then we should return all operations regardless of the span kind |
That means including all spans of old data also (Whose kind is not there in cache)? |
My current approach is leading to errors in unit test of
This is probably because
The only problem is that, during prefilling 6*NumberOfOperations Get Queries will be called. Please review this approach @yurishkuro and I think we need to discuss about autoCreation of new index or should we skip the creation of any new index and use the function given above? |
@yurishkuro I finally got rid of migration and now I think its ready for review! Please ignore my previous comments. The current commit has no linkage them! |
|
@yurishkuro This PR is ready to review, I have added dual lookups and backward compatibility tests in this PR. |
} | ||
err := writer.writeSpanWithOldIndex(&oldSpan) | ||
require.NoError(t, err) | ||
traces, err := reader.FindTraces(context.Background(), &spanstore.TraceQueryParameters{ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
not sure I follow this test. What does FindTraces have to do with span kind in the operations retrieval? Also, backwards compatibility test only makes sense when it is executed against old and new code.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We have changed the key but we need to make sure that traces are also fetched from old key when dual lookup is turned on. Please stress on a fact that operation key is used in getting traces also along with filling in cache, If you will look at this code, we are first writing span with old key and then testing whether it is able to fetch traces associated with that key (please see L42)
} | ||
*/ | ||
// The uint64 value is the expiry time of operation | ||
operations map[string]map[model.SpanKind]map[string]uint64 |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
to clarify, CacheStore is used to avoid expensive scans when loading services and operations, correct? In other words, it's all in-memory structure. In this case, why can we not change just the value of the map to be a combo {kind, expiration}
instead of changing the structure? When loading, scanning everything for a give service is still going to be negligible amount of data.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can't understand this! Are you saying to keep these structures?
services map[string]uint64 // Already in the cache
operations map[string][string]kind
type kind struct {
kind SpanKind
expiry uint64
}
If yes, then how to handle when query is to fetch all operations for a service and kind? Should we iterate all operations and skip those operations which are not of the required kind? (We are using a similar approach currently, i.e iteralting for all kinds and skipping unrequired kinds but this was justified because max kinds can be 6 but number of operations aren't defined, so will this option viable?)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, this structure.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
So iterating all operations and skipping not required kinds will be right?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
While approaching towards this, I am leading to a conclusion that this approach will lead to the same problem that spans with same operation and service name but different kind will end up in overriding of data. So I don't think that this structure is going to be a correct approach! Rather I could think of only 3D map a viable option. So should we move forward with 3D map or can we have a better idea?
…er (#6575) ## Which problem is this PR solving? Comment: #6376 (comment) ## Description of the changes - Cache was directly contacting the db to prefill itself which is not a good way, now this responsibility is given to reader to read from badger and fill the cache. ## How was this change tested? - Unit and e2e tests ## Checklist - [x] I have read https://github.com/jaegertracing/jaeger/blob/master/CONTRIBUTING_GUIDELINES.md - [x] I have signed all commits - [x] I have added unit tests for the new functionality - [x] I have run lint and test steps successfully - for `jaeger`: `make lint test` - for `jaeger-ui`: `npm run lint` and `npm run test` --------- Signed-off-by: Manik2708 <[email protected]>
@yurishkuro A humble reminder to review this PR! |
Signed-off-by: Manik2708 <[email protected]>
} | ||
|
||
// This method is to test backward compatibility for old index key | ||
func (w *SpanWriter) writeSpan(span *model.Span, writeOldIndex bool) error { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Not sure about this approach. But my aim was to lower the diff!
|
||
store *badger.DB |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can't find any use of store in Cache when the responsibility to fill cache is given to reader
"jaeger.badger.dualLookUp", | ||
featuregate.StageBeta, // enabed by default | ||
featuregate.WithRegisterFromVersion("v2.2.0"), | ||
featuregate.WithRegisterToVersion("v2.5.0"), |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I am confused about the two options:
- Versions used
- Whether it should be the issue which should be linked or pull request, as issue is not talking about this change directly!
Signed-off-by: Manik Mehta <[email protected]>
Signed-off-by: Manik2708 <[email protected]>
@yurishkuro Sorry for disturbance, but can you please review this PR and resolve the doubts? |
/gemini review |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Code Review
The pull request introduces SpanKind support for Badger, which is a valuable enhancement. The changes seem well-structured and include a backward compatibility test. However, there are a few areas that could benefit from further review and refinement.
Summary of Findings
Merge Readiness
The pull request appears to be in good shape overall, but I recommend addressing the comments provided below before merging. I am unable to directly approve this pull request, so please ensure that other reviewers also examine the changes and provide their approval before proceeding with the merge.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
PR Overview
This PR adds support for SpanKind in Badger storage by updating index keys, reader/writer logic, the cache structure, and integration tests.
- Introduces new type mappings and conversion functions for SpanKind.
- Updates factory, reader, writer, and cache layers to create and query indexes that incorporate span kind.
- Adds and adjusts unit and integration tests to verify backward compatibility and new functionality.
Reviewed Changes
File | Description |
---|---|
internal/storage/v1/badger/spanstore/backward_compatibility_test.go | Adds tests to ensure backward compatibility after the index changes. |
internal/storage/v1/badger/spanstore/kind.go | Introduces new type and mapping functions for span kind conversion. |
internal/storage/v1/badger/factory.go | Registers a new feature gate and updates cache initialization and reader construction. |
internal/storage/v1/badger/spanstore/reader.go | Updates TraceReader to accept a dual lookup flag and prefill operations by span kind. |
internal/storage/v1/badger/spanstore/rw_internal_test.go | Adjusts tests to use the new TraceReader and CacheStore signatures. |
internal/storage/v1/badger/spanstore/writer.go | Modifies index key creation to incorporate SpanKind for operations. |
internal/storage/v1/badger/spanstore/cache.go | Updates cache structure for operations to be keyed by span kind. |
internal/storage/integration/badgerstore_test.go, cmd/jaeger/internal/integration/badger_test.go | Remove legacy flags and update integration tests to reflect SpanKind support. |
Copilot reviewed 10 out of 10 changed files in this pull request and generated 1 comment.
Comments suppressed due to low confidence (1)
internal/storage/v1/badger/spanstore/reader.go:60
- [nitpick] Consider renaming 'dualLookUp' to 'dualLookup' to follow common naming conventions.
dualLookUp bool
Co-authored-by: Copilot <[email protected]> Signed-off-by: Manik Mehta <[email protected]>
Signed-off-by: Manik2708 <[email protected]>
Which problem is this PR solving?
Description of the changes
How was this change tested?
Checklist
jaeger
:make lint test
jaeger-ui
:npm run lint
andnpm run test