GODRIVER-2101 Direct read/write retries to another mongos if possible #1358
Conversation
This file contains 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
4 tasks
prestonvasquez
requested review from
qingyang-hu and
matthewdale
and removed request for
a team
prestonvasquez
commented
Aug 25, 2023
| return nil, err | ||
| } | ||
|
|
||
| filteredServers := filterDeprioritizedServers(selectedServers, oss.deprioritizedServers) |
There was a problem hiding this comment.
This method was designed to add more filters if the need arrises in the future.
qingyang-hu
reviewed
Aug 28, 2023
qingyang-hu
reviewed
Aug 28, 2023
prestonvasquez
had a problem deploying
to
api-report
GitHub Actions
Failure
prestonvasquez
had a problem deploying
to
api-report
GitHub Actions
Failure
matthewdale
previously approved these changes
Sep 11, 2023
Comment on lines
+293
to
+296
| // Note that setting this value greater than 2 will result in false | ||
| // negatives. The current specification does not account for CSOT, which | ||
| // might allow for an "inifinite" number of retries over a period of time. | ||
| // Because of this, we only track the "previous server". |
There was a problem hiding this comment.
Is there a task for updating the retryable reads/writes "deprioritized mongos" behavior to account for multiple retries (i.e. CSOT)? The vast majority of sharded clusters have >2 mongos nodes, so that seems like a questionably useful feature for drivers that support CSOT.
There was a problem hiding this comment.
For now, there is no task to do this. Here are a couple of reasons from discussions with @comandeo:
- We do not want this new mechanism to replace SDAM/interfere with SDAM too much.
- We believe that mongos may recover from the error fast enough, and there is no reason to exclude ones that failed earlier
- It is rather a rare occasion that multiple mongoses fail with retryable errors. This looks like a network issue, and this is handled by SDAM
qingyang-hu
previously approved these changes
Sep 11, 2023
prestonvasquez
dismissed stale reviews from qingyang-hu and matthewdale
via
943ecbe
prestonvasquez
temporarily deployed
to
api-report
GitHub Actions
Inactive
API Change ReportNo changes found! |
qingyang-hu
approved these changes
Sep 12, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
GODRIVER-2101
Summary
When possible, deprioritize failed mongos during retry attempts.
Background & Motivation
Our current retry logic for sharded clusters can lead to an operation that failed with a retryable error being retried on the same mongos.