[
https://issues.apache.org/jira/browse/HADOOP-17461?focusedWorklogId=782504&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-782504
]
ASF GitHub Bot logged work on HADOOP-17461:
-------------------------------------------
Author: ASF GitHub Bot
Created on: 17/Jun/22 18:31
Start Date: 17/Jun/22 18:31
Worklog Time Spent: 10m
Work Description: steveloughran commented on PR #4352:
URL: https://github.com/apache/hadoop/pull/4352#issuecomment-1159135030
we need a single IOStatisticsContext for all FS instances, so that a task
reading from one fs and writing to another would have the stats updated from
both actions.
And it needs to be in hadoop-common, just like the common audit context, so
that code can compile and link against it even without having hadoop-aws or
hadoop-azure on the classpath.
Making the context a weak ref map ensures that GCs will trigger cleanup.
This would lose stats, but not while any stream was active, *or if some code
picked up a reference to the thread stats before executing work. That is what I
plan to do in spark; we will grab that ref before starting the work, and after
it is finished, take a snapshot of it. Oh, and reset the values before work
starts -we need that context there too.
Issue Time Tracking
-------------------
Worklog Id: (was: 782504)
Time Spent: 1h 20m (was: 1h 10m)
> Add thread-level IOStatistics Context
> -------------------------------------
>
> Key: HADOOP-17461
> URL: https://issues.apache.org/jira/browse/HADOOP-17461
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs, fs/azure, fs/s3
> Affects Versions: 3.3.1
> Reporter: Steve Loughran
> Assignee: Mehakmeet Singh
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> For effective reporting of the iostatistics of individual worker threads, we
> need a thread-level context which IO components update.
> * this contact needs to be passed in two background thread forming work on
> behalf of a task.
> * IO Components (streams, iterators, filesystems) need to update this context
> statistics as they perform work
> * Without double counting anything.
> I imagine a ThreadLocal IOStatisticContext which will be updated in the
> FileSystem API Calls. This context MUST be passed into the background threads
> used by a task, so that IO is correctly aggregated.
> I don't want streams, listIterators &c to do the updating as there is more
> risk of double counting. However, we need to see their statistics if we want
> to know things like "bytes discarded in backwards seeks". And I don't want to
> be updating a shared context object on every read() call.
> If all we want is store IO (HEAD, GET, DELETE, list performance etc) then the
> FS is sufficient.
> If we do want the stream-specific detail, then I propose
> * caching the context in the constructor
> * updating it only in close() or unbuffer() (as we do from S3AInputStream to
> S3AInstrumenation)
> * excluding those we know the FS already collects.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
