Add new SamzaApplicationMaster metric to track allocated containers buffered in AM #1677
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
…y RM and buffered in AM
jia-gao
changed the title
shanthoosh
reviewed
Jul 18, 2023
samza-yarn/src/main/java/org/apache/samza/job/yarn/YarnClusterResourceManager.java
Show resolved
Hide resolved
shanthoosh
reviewed
Jul 18, 2023
samza-yarn/src/main/scala/org/apache/samza/job/yarn/SamzaAppMasterMetrics.scala
Show resolved
Hide resolved
shanthoosh
approved these changes
Jul 18, 2023
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Outdated suggestions cannot be applied.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.
LISAMZA-28234
Symptom:
We have observed that Samza YARN AM requests 2-3x more containers than needed during job startup. Can cause temporary resource exhaustion in cluster when large jobs are launched, causing unrelated deployments to timeout/fail.
Change:
This PR added a Samza AM metric to monitor how many containers are allocated by RM and buffered in AM
The metric increments when AM got an allocated container from RM and add it to the buffer. The metric decrements when AM released the overallocated containers (after all processors started).
By observing this metric, we can know how many containers were allocated to a job during its startup and how long before the overallocated ones got released.