[
https://issues.apache.org/jira/browse/FLINK-27155?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526298#comment-17526298
]
Roman Khachatryan commented on FLINK-27155:
-------------------------------------------
I think Yun Tang means waiting for all task of a TM to switch from INITIALIZING
to RUNNING.
I think it could be problematic, because the set of tasks is dynamic.
Besides that, because the local space can be limited, and changelog can be
large, it's better to clean up the cache earlier.
Probably, we should combine reference counting and timeouts.
periodic-materialize.interval can still be less than the recovery time.
As for serializing file accesses, I'm afraid that without it the ticket can
introduce regression. Furthermore, the need to serialize can affect the design
of cache. So I'd at least design this at once.
> Reduce multiple reads to the same Changelog file in the same taskmanager
> during restore
> ---------------------------------------------------------------------------------------
>
> Key: FLINK-27155
> URL: https://issues.apache.org/jira/browse/FLINK-27155
> Project: Flink
> Issue Type: Sub-task
> Components: Runtime / State Backends
> Reporter: Feifan Wang
> Assignee: Feifan Wang
> Priority: Major
> Fix For: 1.16.0
>
>
> h3. Background
> In the current implementation, State changes of different operators in the
> same taskmanager may be written to the same changelog file, which effectively
> reduces the number of files and requests to DFS.
> But on the other hand, the current implementation also reads the same
> changelog file multiple times on recovery. More specifically, the number of
> times the same changelog file is accessed is related to the number of
> ChangeSets contained in it. And since each read needs to skip the preceding
> bytes, this network traffic is also wasted.
> The result is a lot of unnecessary request to DFS when there are multiple
> slots and keyed state in the same taskmanager.
> h3. Proposal
> We can reduce multiple reads to the same changelog file in the same
> taskmanager during restore.
> One possible approach is to read the changelog file all at once and cache it
> in memory or local file for a period of time when reading the changelog file.
> I think this could be a subtask of [v2 FLIP-158: Generalized incremental
> checkpoints|https://issues.apache.org/jira/browse/FLINK-25842] .
> Hi [~ym] , [~roman] how do you think about ?
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
