You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ozone.apache.org by "Siyao Meng (Jira)" <ji...@apache.org> on 2023/09/02 15:23:00 UTC
[jira] [Updated] (HDDS-9233) Investigate whether listStatus() is correctly iterating cache
[ https://issues.apache.org/jira/browse/HDDS-9233?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Siyao Meng updated HDDS-9233:
-----------------------------
Description:
It has come to my attention that the cleanup() method in TestOzoneFileSystem (in CI integration (filesystem)), which is triggered after EACH test method is incredibly flaky.
However, when I add a call to wait for double buffer flush, it no longer seems flaky:
https://github.com/apache/ozone/pull/5175/commits/72c297d24a27bd6f334979f70a7a927fc03d8114
I suspect ofs/o3fs listStatus("/") is not correctly iterating cache. Otherwise it should be strongly consistent -- any file deletion should be immediately reflected to this or other clients.
was:
It has come to my notice that the cleanup() method in TestOzoneFileSystem, which is triggered after EACH test method is incredibly flaky.
However, when I add a call to wait for double buffer flush, it no longer seems flaky:
https://github.com/apache/ozone/pull/5175/commits/72c297d24a27bd6f334979f70a7a927fc03d8114
I suspect ofs/o3fs listStatus("/") is not correctly iterating cache. Otherwise it should be strongly consistent -- any file deletion should be immediately reflected to this or other clients.
> Investigate whether listStatus() is correctly iterating cache
> -------------------------------------------------------------
>
> Key: HDDS-9233
> URL: https://issues.apache.org/jira/browse/HDDS-9233
> Project: Apache Ozone
> Issue Type: Task
> Reporter: Siyao Meng
> Assignee: Devesh Kumar Singh
> Priority: Critical
>
> It has come to my attention that the cleanup() method in TestOzoneFileSystem (in CI integration (filesystem)), which is triggered after EACH test method is incredibly flaky.
> However, when I add a call to wait for double buffer flush, it no longer seems flaky:
> https://github.com/apache/ozone/pull/5175/commits/72c297d24a27bd6f334979f70a7a927fc03d8114
> I suspect ofs/o3fs listStatus("/") is not correctly iterating cache. Otherwise it should be strongly consistent -- any file deletion should be immediately reflected to this or other clients.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org