Skip to content

[SPARK-22907]Clean broadcast garbage when IOException occurred in MetadataFetch - #20090

Closed
liupc wants to merge 2 commits into
apache:masterfrom
liupc:Clean-broadcast-garbage-when-ioexception-occurred-in-metadataFetch
Closed

[SPARK-22907]Clean broadcast garbage when IOException occurred in MetadataFetch#20090
liupc wants to merge 2 commits into
apache:masterfrom
liupc:Clean-broadcast-garbage-when-ioexception-occurred-in-metadataFetch

Conversation

@liupc

@liupcliupc commented Dec 27, 2017

Copy link
Copy Markdown

What changes were proposed in this pull request?

When fetching mapStatuses, IOException is not handled in MapOutputTracker.deserializeMapStatuses which might cause old broadcast not cleaned up. In task retrying, when task was scheduled to the executor where old broadcast was stored, An "Block is already present in the MemoryStore" exception would be thrown.

This PR will clean garbage broadcast when IOException occurred in MetadataFetch to avoid
repeated "Block is already present in the MemoryStore" caused by the uncleaned broadcast garbage

details see: https://issues.apache.org/jira/browse/SPARK-22907

(Please fill in changes proposed in this fix)

How was this patch tested?

manual
(Please explain how this patch was tested. E.g. unit tests, integration tests, manual tests)
(If this patch involves UI changes, please attach a screenshot; otherwise, remove this)

Please review http://spark.apache.org/contributing.html before opening a pull request.

liupengcheng added 2 commits December 27, 2017 17:36
Summary: Ref T10399
Test Plan: n/a
Reviewers: liushaohui, peng.zhang, zhoukang, chenfan
Subscribers: cloud-computing
Maniphest Tasks: T10399
Differential Revision: https://phabricator.d.xiaomi.net/D79279
@squito

Copy link
Copy Markdown
Contributor

Thanks for submitting this @liupc. The jira description makes sense, though I need to take a closer look at the code to confirm things. But at least two things need to be fixed before this can be merged:

  1. can you clean up the PR description?

  2. can you add a test case for this?

@jiangxb1987

Copy link
Copy Markdown
Contributor

ping @liupc

@AmplabJenkins

Copy link
Copy Markdown

Can one of the admins verify this patch?

@HyukjinKwon

Copy link
Copy Markdown
Member

ping @liupc

@HyukjinKwonHyukjinKwon mentioned this pull request Jul 16, 2018
zifeif2 pushed a commit to zifeif2/spark that referenced this pull request Nov 22, 2025
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants

@liupc@squito@jiangxb1987@AmplabJenkins@HyukjinKwon