Cherry-pick Gray failure detection and recovery to release 6.3 #5249
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
Contributor
AWS CodeBuild CI Report
|
Contributor
AWS CodeBuild CI Report
|
Contributor
AWS CodeBuild CI Report
|
Contributor
AWS CodeBuild CI Report
|
halfprice
force-pushed
the
zhewu/gray-failure-6.3
branch
from
ed7a75d to
c4b0814
Compare
Contributor
AWS CodeBuild CI Report
|
Contributor
AWS CodeBuild CI Report
|
…is change is only inside the worker.
… of ClusterControllerFullInterface
halfprice
force-pushed
the
zhewu/gray-failure-6.3
branch
from
c4b0814 to
f658d45
Compare
Contributor
AWS CodeBuild CI Report
|
halfprice
force-pushed
the
zhewu/gray-failure-6.3
branch
from
f658d45 to
590feae
Compare
halfprice
changed the title
Contributor
AWS CodeBuild CI Report
|
jzhou77
approved these changes
Aug 16, 2021
RenxuanW
approved these changes
Aug 19, 2021
xumengpanda
reviewed
Aug 19, 2021
|
|
||
| auto& health = workerHealth[workerAddress]; | ||
|
|
||
| // First, remove any degraded peers recorded in the `workerHealth`, but aren't in the incoming request. These |
Contributor
There was a problem hiding this comment.
the format looks funny.
likely tab vs space.
can you run clang-format to the change?
xumengpanda
reviewed
Aug 19, 2021
| // degraded since A is already considered as degraded. | ||
| std::unordered_set<NetworkAddress> currentDegradedServers; | ||
| for (const auto& [complainerCount, badServer] : count2DegradedPeer) { | ||
| for (const auto& complainer : degradedLinkDst2Src[badServer]) { |
Contributor
There was a problem hiding this comment.
so if a server is considered as degraded and excluded, all peers of the degraded server won't be considered as degraded.
is that a correct summary?
xumengpanda
reviewed
Aug 19, 2021
| } | ||
| } else { | ||
| self->excludedDegradedServers.clear(); | ||
| TraceEvent("DegradedServerDetectedAndSuggestRecovery"); |
Contributor
There was a problem hiding this comment.
can we make the event more straightforward to SRE on which server to exclude before trigger recovery?
e.g., add a field.
.detail("Hint", "Exclude based on ClusterControllerHealthMonitor event and manually trigger recovery")
xumengpanda
reviewed
Aug 19, 2021
| init( CC_MIN_DEGRADATION_INTERVAL, 120.0 ); | ||
| init( CC_DEGRADED_PEER_DEGREE_TO_EXCLUDE, 3 ); | ||
| init( CC_MAX_EXCLUSION_DUE_TO_HEALTH, 2 ); | ||
| init( CC_HEALTH_TRIGGER_RECOVERY, false ); |
Contributor
There was a problem hiding this comment.
would simulation pass if we buggify this knob to true?
xumengpanda
reviewed
Aug 19, 2021
| init( MIN_DELAY_CC_WORST_FIT_CANDIDACY_SECONDS, 10.0 ); | ||
| init( MAX_DELAY_CC_WORST_FIT_CANDIDACY_SECONDS, 30.0 ); | ||
| init( DBINFO_FAILED_DELAY, 1.0 ); | ||
| init( ENABLE_WORKER_HEALTH_MONITOR, false ); |
Contributor
There was a problem hiding this comment.
since SRE may toggle some of the knobs, it would be good to add if( randomize && BUGGIFY ) to test different values of those knobs.
5 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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.
This PR cherry-picks the new single cluster gray failure detection and recovery mechanism to release 7.0 branch.
This feature is guarded by knobs and currently turned off by default.
Tests performed on this PR:
Code-Reviewer Section
The general guidelines can be found here.
Please check each of the following things and check all boxes before accepting a PR.
For Release-Branches
If this PR is made against a release-branch, please also check the following:
release-branchormasterif this is the youngest branch)