[ 
https://issues.apache.org/jira/browse/ASTERIXDB-3226?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17749441#comment-17749441
 ] 

ASF subversion and git services commented on ASTERIXDB-3226:
------------------------------------------------------------

Commit e64d04628930c3aac2d4b3d10a098f35e9ca0464 in asterixdb's branch 
refs/heads/master from Murtadha Makki Al Hubail
[ https://gitbox.apache.org/repos/asf?p=asterixdb.git;h=e64d046289 ]

[ASTERIXDB-3226][STO] Warn on missing primary index during flush

- user model changes: no
- storage format changes: no
- interface changes: no

Details:

- To account for leaked secondary files, warn rather than halting
  on missing primary index during flush to allow recovery to complete.
  Eventually, the leaked files will be deleted during global recovery.
- Fix log message.

Change-Id: I7858a67cfc97fb1fde01c463a256083952d51543
Reviewed-on: https://asterix-gerrit.ics.uci.edu/c/asterixdb/+/17647
Reviewed-by: Michael Blow <mb...@apache.org>
Integration-Tests: Jenkins <jenk...@fulliautomatix.ics.uci.edu>
Tested-by: Jenkins <jenk...@fulliautomatix.ics.uci.edu>


> Crash loop on leaked secondary index files
> ------------------------------------------
>
>                 Key: ASTERIXDB-3226
>                 URL: https://issues.apache.org/jira/browse/ASTERIXDB-3226
>             Project: Apache AsterixDB
>          Issue Type: Bug
>          Components: STO - Storage
>    Affects Versions: 0.9.8
>            Reporter: Murtadha Makki Al Hubail
>            Assignee: Murtadha Makki Al Hubail
>            Priority: Major
>              Labels: triaged
>             Fix For: 0.9.9
>
>
> When some secondary index files are leaked after its primary is dropped, an 
> NC will be in a crash loop after a restart due to the startup checkpoint task 
> attempting to flush all registered indexes and we halt on missing primary 
> during flush. Since there are scenarios where secondary index files can be 
> leaked, we should warn rather than halt. Eventually, when all NCs complete 
> their startup, the global storage cleanup will remove such indexes.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

Reply via email to