Update default maxSegmentsInNodeLoadingQueue#11540
Merged
suneet-s merged 3 commits intoapache:masterfrom Aug 5, 2021
Merged
Conversation
Update the default maxSegmentsInNodeLoadingQueue from 0 (unbounded) to 100. An unbounded maxSegmentsInNodeLoadingQueue can cause cluster instability. Since this is the default druid operators need to run into this instability and then look through the docs to see that the recommended value for a large cluster is 1000. This change makes it so the default will prevent clusters from falling over as they grow over time.
Contributor
Author
|
Added |
Contributor
|
LGTM. |
kfaraz
requested changes
Aug 4, 2021
| // updated to Jackson 2.9 it could be changed, see https://github.com/apache/druid/issues/7152 | ||
| @JsonProperty("killPendingSegmentsSkipList") Object dataSourcesToNotKillStalePendingSegmentsIn, | ||
| @JsonProperty("maxSegmentsInNodeLoadingQueue") int maxSegmentsInNodeLoadingQueue, | ||
| @JsonProperty(value = "maxSegmentsInNodeLoadingQueue") @Nullable Integer maxSegmentsInNodeLoadingQueue, |
Contributor
There was a problem hiding this comment.
Nit: Is the value = part required?
| replicateAfterLoadTimeout == null ? DEFAULT_REPLICATE_AFTER_LOAD_TIMEOUT : replicateAfterLoadTimeout, | ||
| maxNonPrimaryReplicantsToLoad == null ? DEFAULT_MAX_NON_PRIMARY_REPLICANTS_TO_LOAD : maxNonPrimaryReplicantsToLoad | ||
| maxNonPrimaryReplicantsToLoad == null | ||
| ? DEFAULT_MAX_NON_PRIMARY_REPLICANTS_TO_LOAD |
Contributor
Author
There was a problem hiding this comment.
I thought so too, but I used intelliJ's auto-format and this is what it came up with 🤷
Contributor
|
LGTM |
kfaraz
approved these changes
Aug 4, 2021
jihoonson
approved these changes
Aug 5, 2021
maytasm
approved these changes
Aug 5, 2021
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
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.
Description
Update the default maxSegmentsInNodeLoadingQueue from 0 (unbounded) to 100.
An unbounded maxSegmentsInNodeLoadingQueue can cause cluster instability.
Since this is the default druid operators need to run into this instability
and then look through the docs to see that the recommended value for a large
cluster is 1000. This change makes it so the default will prevent clusters
from falling over as they grow over time.
This PR has: