REP-5318 Retry on change stream failures. #52
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
FGasper
force-pushed
the
felipe_retry_change_stream
branch
3 times, most recently
from
8231711 to
03853c7
Compare
FGasper
changed the title
FGasper
force-pushed
the
felipe_retry_change_stream
branch
from
03853c7 to
1348a60
Compare
tdq45gj
requested changes
Nov 25, 2024
Collaborator
tdq45gj
left a comment
tdq45gj
left a comment
There was a problem hiding this comment.
Looks good. I think we need to call IterationSuccess when iterating the change stream.
internal/verifier/change_stream.go
Outdated
| // StartChangeStream starts the change stream. | ||
| func (verifier *Verifier) StartChangeStream(ctx context.Context) error { | ||
| // Result seems a bit simpler than messing with 2 separate channels. | ||
| resultChan := make(chan mo.Result[primitive.Timestamp]) |
Collaborator
There was a problem hiding this comment.
I'm a bit confused. Could you clarify what result does this resultChan expect?
Collaborator
Author
There was a problem hiding this comment.
Result holds a “payload, or error”. It’s a common pattern in Rust.
The alternative would be 2 separate channels then a select on those.
Collaborator
There was a problem hiding this comment.
Does it hold the result from createChangeStream? Could we use a more descriptive name such as createChangeStreamResultChan?
Collaborator
Author
There was a problem hiding this comment.
It’s the first change stream creation’s result. I’ve reshuffled the code a bit to clarify that.
Previously this function actually returned the most recent cluster time rather than creating a new cluster time. This changeset makes it now create a new time instead. This simplifies the change stream’s handling of the writesOff timestamp because the writesOff timestamp is separate from any change event.
FGasper
force-pushed
the
felipe_retry_change_stream
branch
from
0f074f1 to
1547534
Compare
This entails a small refactor of the change stream code so that the change stream’s creation and iteration both happen under a retryer.
FGasper
force-pushed
the
felipe_retry_change_stream
branch
from
1547534 to
f4fcc05
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
REVIEW NOTE
It probably makes sense to review this PR’s 2 commits separately.
This is necessary for testing migration-verifier’s change stream in mongosync’s nondeterministic tests. It entails a small refactor of the change stream code so that both the change stream’s creation and iteration happen under a retryer.
Additionally, this makes GetNewClusterTime() return an actual new cluster time rather than returning the current one. This simplifies reasoning about the change stream’s handling of the writesOff timestamp.