-
Notifications
You must be signed in to change notification settings - Fork 9.1k
HADOOP-17596. ABFS: Change default Readahead Queue Depth from num(processors) to const #2795
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
|
💔 -1 overall
This message was automatically generated. |
|
🎊 +1 overall
This message was automatically generated. |
|
TEST RESULTS HNS Account Location: East US 2 |
|
🎊 +1 overall
This message was automatically generated. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
+1
...s/hadoop-azure/src/test/java/org/apache/hadoop/fs/azurebfs/services/TestAbfsInputStream.java
Outdated
Show resolved
Hide resolved
|
When there are not many parallel FIS reads, this (possible) reduction in read ahead parallelism will affect the read throughput? Specially Spark kind of? From HBase standpoint this is acceptable. |
|
🎊 +1 overall
This message was automatically generated. |
|
@anoopsjohn ==> wrt your comment on "will this hamper spark jobs".
|
|
Thanks @vinaysbadami . Ya speculative part I agree. Points #2 and #3 are very important. You can think whether can include this in abfs.md even. Will be good.
No I dont think so. This config is cluster level and all input streams opened in this will use this value. |
We can follow up in a jira. Will be nice to have as a feature support from Driver side. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
+1
|
Thanks @anoopsjohn for review |
Keeping both branches in sync is essential for cherrypicking future work |
|
@steveloughran , I updated the Jira.
Yes, @sumangala-patki is working on backport and testing it for 3.2 and 3.3 branch. |
|
@surendralilhore thanks for updating the JIRA |
…cessors) to const (apache#2795) . Contributed by Sumangala Patki. (cherry picked from commit 76d92eb)
…cessors) to const (apache#2795) . Contributed by Sumangala Patki.
…cessors) to const (#3106) * HADOOP-17596. ABFS: Change default Readahead Queue Depth from num(processors) to const (#2795) . Contributed by Sumangala Patki. (cherry picked from commit 76d92eb)
…cessors) to const (apache#2795) . Contributed by Sumangala Patki.
The default value of readahead queue depth is currently set to the number of available processors. However, this can result in one inputstream instance consuming more processor time. To ensure equal thread allocation during read for all inputstreams created in a session, we change the default readahead queue depth to a constant (2).