Improved redirection error handling #857
Merged
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
nihohit
approved these changes
Jun 6, 2023
nihohit
reviewed
Jun 13, 2023
This PR adds proper `ASK` redirection for the async cluster implementation, so rather than simply rebuilding slots, which may not be an appropriate response to `ASK` anyway, it issues the relevant `ASKING` command to the redirected node and issues the query. A `Redirect` enum has been added for better handling of both redirect scenarios. The async cluster implementation has also been tweaked. 1) Instead of immediately refreshing slots, the redirected node will be queried, and slots will be refreshed after the redirected request has been made. The same approach has not been taken on the sync implementation for now because doing so would require greater refactoring than is appropriate here. 2) To support 1), async-cluster requests have been refactored to try and create a connection to the redirected node if it doesn't already exist in the connection-map (which is likely, of course, if a new node has been added to the cluster). A couple existing tests have also been renamed for clarity.
altanozlu
pushed a commit
to altanozlu/redis-rs
that referenced
this pull request
Aug 16, 2023
This PR adds proper `ASK` redirection for the async cluster implementation, so rather than simply rebuilding slots, which may not be an appropriate response to `ASK` anyway, it issues the relevant `ASKING` command to the redirected node and issues the query. A `Redirect` enum has been added for better handling of both redirect scenarios. The async cluster implementation has also been tweaked as follows: 1) Instead of immediately refreshing slots, the redirected node will be queried, and slots will be "scheduled" for refresh after the redirected request (and other "in-flight requests") has been made. The same approach has not been taken on the sync implementation for now because doing so would require greater refactoring than is appropriate here. 2) To support 1), async-cluster requests have been refactored to potentially create a connection to the appropriate address if an existing one cannot be found or obtained (which is likely, of course, if a new node has been added to the cluster). This matches existing functionality in the sync implementation. A couple existing tests have also been renamed for clarity.
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.
This PR adds proper
ASKredirection for the async cluster implementation, so rather than simply rebuilding slots, which may not be an appropriate response toASKanyway, it issues the relevantASKINGcommand to the redirected node and issues the query.A
Redirectenum has been added for better handling of both redirect scenarios.The async cluster
MOVEDimplementation has also been tweaked. Instead of immediately refreshing slots, the redirected node will be queried, and slots will be refreshed after the redirected request has been made. The same approach has not been taken on the sync implementation for now because doing so would require greater refactoring than is appropriate here.