[DPE-3591] Fix shared buffers validation #361
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
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #361 +/- ##
==========================================
- Coverage 81.33% 81.25% -0.08%
==========================================
Files 10 10
Lines 2207 2235 +28
Branches 353 360 +7
==========================================
+ Hits 1795 1816 +21
- Misses 343 347 +4
- Partials 69 72 +3 ☔ View full report in Codecov by Sentry. |
…e cache size value Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
dragomirp
approved these changes
Feb 16, 2024
…uffers-validation Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
Signed-off-by: Marcelo Henrique Neppel <[email protected]>
…uffers-validation Signed-off-by: Marcelo Henrique Neppel <[email protected]>
* Reenable landscape tests * Try GH runners * Back on self-hosted runner * Don't check for primary if peer didn't join yet * Don't recheck missing primary ip when looking for primary endpoint * Catch list user exception * Catch Psycopg2 error when trying to set config * Try to preflight connection on config validation * Don't retry db connection check * Set the primary endpoint before starting the primary * Revert "Set the primary endpoint before starting the primary" This reverts commit 8184dfa. * Wait for peer before starting primary * Switch to GH runners * Recheck DB connection * Don't check peer on primary startup * Try to skip primary IP validation if not all cluster IPs seem to be available * Comment * Review comments * Unit tests
BON4
pushed a commit
to BON4/postgresql-operator
that referenced
this pull request
Apr 23, 2024
* Fix shared buffers validation Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Consider shared buffers config option value when calculating effective cache size value Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Fix PostgreSQL restart not being called Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Add and update unit test Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Update unit test Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Increase test timeout Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Increase test timeout Signed-off-by: Marcelo Henrique Neppel <[email protected]> * Switch to GH runners * Increase stanza timeout * [MISC] Check for peer relation when getting the primary endpoint (canonical#366) * Reenable landscape tests * Try GH runners * Back on self-hosted runner * Don't check for primary if peer didn't join yet * Don't recheck missing primary ip when looking for primary endpoint * Catch list user exception * Catch Psycopg2 error when trying to set config * Try to preflight connection on config validation * Don't retry db connection check * Set the primary endpoint before starting the primary * Revert "Set the primary endpoint before starting the primary" This reverts commit 8184dfa. * Wait for peer before starting primary * Switch to GH runners * Recheck DB connection * Don't check peer on primary startup * Try to skip primary IP validation if not all cluster IPs seem to be available * Comment * Review comments * Unit tests --------- Signed-off-by: Marcelo Henrique Neppel <[email protected]> Co-authored-by: Dragomir Penev <[email protected]> Co-authored-by: Dragomir Penev <[email protected]>
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.
Issue
The validation for the
shared_buffersparameter was not correct. The calculus of the maximum allowed shared buffers value was not correctly converting the available memory for that to the same unit as theshared_buffersparameter that the user informs.Also, the
effective_cache_sizeparameter did not consider the user-providedshared_buffersvalue.One last issue: sometimes, the restart of the PostgreSQL database was not triggered because Patroni hadn't finished reloading the configuration (even after 10 seconds).
Solution
Fix the calculus of the maximum allowed shared buffers value to be in the same unit as the user-provided value.
Considered the user-provided shared buffers value for the effective cache size parameter calculus.
An improved retry mechanism was added to check whether a restart is needed after a configuration change.
Unit tests for the library file are being updated to the K8S charm, which owns the library.