perf(kubernetes): Reduce memory allocation during caching cycles #3736
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
Also simplify the logic in the invertRelationships test
The caching logic for the kubernetes v2 provider allocates a lot of short-term memory during each caching cycle, putting pressure on the garbage collector, in some cases exceeding the garbage collection overhead. A significant contributor is the logic to compute relationships between kubernetes objects. We currently create a CacheData object to hold each relationship and rely on downstream (and inefficient) logic to merge these into one CacheData per object, containing all of its relationships. Improve this by having invertRelationships return one CacheData per object (containing all of its relationships) rather than on CacheData per relationship. Likewise, have getClusterRelationships return a single CacheData for each application, rather than a separate one for each application-object relationship.
|
This should help significantly with spinnaker/spinnaker#3287. |
|
@spinnakerbot cherry-pick 1.14 |
spinnakerbot
pushed a commit
that referenced
this pull request
Jun 4, 2019
* perf(kubernetes): Reduce memory allocation during caching cycles The caching logic for the kubernetes v2 provider allocates a lot of short-term memory during each caching cycle, putting pressure on the garbage collector, in some cases exceeding the garbage collection overhead. A significant contributor is the logic to compute relationships between kubernetes objects. We currently create a CacheData object to hold each relationship and rely on downstream (and inefficient) logic to merge these into one CacheData per object, containing all of its relationships. Improve this by having invertRelationships return one CacheData per object (containing all of its relationships) rather than on CacheData per relationship. Likewise, have getClusterRelationships return a single CacheData for each application, rather than a separate one for each application-object relationship. * test(kubernetes): Add tests to getClusterRelationships Also simplify the logic in the invertRelationships test
|
Cherry pick successful: #3744 |
ezimanyi
pushed a commit
that referenced
this pull request
Jun 4, 2019
…) (#3744) * perf(kubernetes): Reduce memory allocation during caching cycles The caching logic for the kubernetes v2 provider allocates a lot of short-term memory during each caching cycle, putting pressure on the garbage collector, in some cases exceeding the garbage collection overhead. A significant contributor is the logic to compute relationships between kubernetes objects. We currently create a CacheData object to hold each relationship and rely on downstream (and inefficient) logic to merge these into one CacheData per object, containing all of its relationships. Improve this by having invertRelationships return one CacheData per object (containing all of its relationships) rather than on CacheData per relationship. Likewise, have getClusterRelationships return a single CacheData for each application, rather than a separate one for each application-object relationship. * test(kubernetes): Add tests to getClusterRelationships Also simplify the logic in the invertRelationships test
maggieneterval
pushed a commit
that referenced
this pull request
Jun 4, 2019
…) (#3746) * perf(kubernetes): Reduce memory allocation during caching cycles The caching logic for the kubernetes v2 provider allocates a lot of short-term memory during each caching cycle, putting pressure on the garbage collector, in some cases exceeding the garbage collection overhead. A significant contributor is the logic to compute relationships between kubernetes objects. We currently create a CacheData object to hold each relationship and rely on downstream (and inefficient) logic to merge these into one CacheData per object, containing all of its relationships. Improve this by having invertRelationships return one CacheData per object (containing all of its relationships) rather than on CacheData per relationship. Likewise, have getClusterRelationships return a single CacheData for each application, rather than a separate one for each application-object relationship. * test(kubernetes): Add tests to getClusterRelationships Also simplify the logic in the invertRelationships test
benjaminws
added a commit
to DataDog/clouddriver
that referenced
this pull request
Jun 26, 2019
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.
test(kubernetes): Add tests to getClusterRelationships
Also simplify the logic in the invertRelationships test
perf(kubernetes): Reduce memory allocation during caching cycles
The caching logic for the kubernetes v2 provider allocates a lot of short-term memory during each caching cycle, putting pressure on the garbage collector, in some cases exceeding the garbage collection overhead.
A significant contributor is the logic to compute relationships between kubernetes objects. We currently create a CacheData object to hold each relationship and rely on downstream (and inefficient) logic to merge these into one CacheData per object, containing all of its relationships.
Improve this by having invertRelationships return one CacheData per object (containing all of its relationships) rather than on CacheData per relationship. Likewise, have getClusterRelationships return a single CacheData for each application, rather than a separate one for each
application-object relationship.