cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benjamin Roth (JIRA)" <>
Subject [jira] [Resolved] (CASSANDRA-12280) nodetool repair hangs
Date Thu, 18 Aug 2016 08:57:22 GMT


Benjamin Roth resolved CASSANDRA-12280.
       Resolution: Not A Bug
    Reproduced In: 3.7, 3.0.8, 3.9  (was: 3.0.8, 3.7, 3.9)

Seems like this was a configuration and/or kernel issue.

For all those who are interested what I did to fix that issue:

- Set tcp keepalive to these values:
Before the total timeout was at about 15min. 90s as in the article fails faster. Maybe that
helps that things to resume faster causing less hangs.
- otc_coalescing_strategy = DISABLED. Don't know if it had a lot of impact but it had no negative
effects. I should double test that but ... I'm glad it works right now.
- Removed a suspicious host. That was the only host with a different kernel and it seemed
that streams from / to this host hung more often than others.

After these changes I did not recognize any (obvious and noticeable) hangs any more.

> nodetool repair hangs
> ---------------------
>                 Key: CASSANDRA-12280
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>            Reporter: Benjamin Roth
> nodetool repair hangs when repairing a keyspace, does not hang when repairting table/mv
by table/mv.
> Command executed (both variants make it hang):
> nodetool repair likes like dislike_by_source_mv like_by_contact_mv match_valid_mv like_out
dislike match match_by_contact_mv like_valid_mv like_out_by_source_mv
> OR
> nodetool repair likes
> Logs:
> Nodetool output:
> Schema:

This message was sent by Atlassian JIRA

View raw message