allow disabling the default golang database/sql retry behavior #899
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
dvilaverde
commented
Jul 24, 2024
|
|
||
| // here we issue assert that even though we only issued 1 query, that the retries | ||
| // remained on and there were 3 calls to the DB. | ||
| require.Equal(t, 3, srv.handler.queryCount) |
There was a problem hiding this comment.
here we had 3 calls to the server mockHandler because the database/sql package retried 3 times and retries were NOT turned off.
dvilaverde
commented
Jul 24, 2024
|
|
||
| // here we issue assert that even though we only issued 1 query, that the retries | ||
| // remained on and there were 3 calls to the DB. | ||
| require.Equal(t, 1, srv.handler.queryCount) |
There was a problem hiding this comment.
same test as before but the retries were off and now there is only 1 call to the server mockHandler
…mysql-org#894) After upgrading to MariaDB 11.4, the canal module stopped detecting row updates within transactions due to incorrect handling of fake rotate events. MariaDB 11.4 does not set LogPos for certain events, causing these events to be ignored. This fix modifies the handling to consider fake rotate events only for ROTATE_EVENTs with timestamp = 0, aligning with MariaDB and MySQL documentation.
lance6716
approved these changes
Jul 25, 2024
morgo
added a commit
to morgo/go-mysql
that referenced
this pull request
Jul 28, 2024
…ditional-ddls * origin/add-additional-ddls: allow disabling the default golang database/sql retry behavior (go-mysql-org#899)
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Outdated suggestions cannot be applied.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.
We've run into a problem caused by the Golangs
database/sqlpackage default retry behavior (which can't be disabled by the way) when an ErrBadConn from thedatabase/sql/driveris returned by the driver. We're looking to have our application fail fast when an issue occurs reaching the DB, but as of now we're seeing 3 retries, causing our application to hang up longer than we'd like.Seems like others in the community have been asking for a way to disable retries, see golang/go#52886 but the go developers have decided to not allow developers to configure that behavior.
This PR makes use of the Validator interface, along with introducing a new Driver argument called
retriesto determine if ErrBadConn or the go-mysql-org native ErrBadConn should be returned.This takes advantage of the fact that retry won't occur unless the database/sql ErrBadConn is returned. Per go docs:
And to maintain the behavior of discarding bad connection, the Validator interface will return false for isValid(), preventing the bad connection from being cached for reuse in sql.DB.
Thoughts on this are appreciated.