You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user@cassandra.apache.org by "marshall.s.king@virtustream.com" <ma...@virtustream.com> on 2017/10/07 17:27:58 UTC
Got error, removing parent repair session - When doing multiple repair -pr ��� Cassandra 3.x
On cassandra 3.11 were getting an error (see error below) when doing multiple repairs with -pr option. Performing this type of repair worked fine on Cassandra 2.x. My questions are:
Is this a valid operation (nodetool repair -pr .. simultaneously on multiple nodes) to do?
If question one is valid, then has anyone seen this error before and if so what do you do to fix?
The error is:
ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 RepairMessageVerbHandler.java:168 - Got error, removing parent repair session ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 CassandraDaemon.java:228 - Exception in thread Thread[AntiEntropyStage:1,5,main] java.lang.RuntimeException: Parent repair session with id = 2689fa80-a2d5-11e7-9a55-b1319b7b5400 has failed.
Thanks, Marshall
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
For additional commands, e-mail: user-help@cassandra.apache.org
Re: RE: Got error, removing parent repair session - When doing multiple repair -pr �������� Cassandra 3.x
Posted by "marshall.s.king@virtustream.com" <ma...@virtustream.com>.
Thanks Thomas for clarifying.
After reading all the threads I see some possible work arounds for this issue. I'm still not exactly sure what the recommended work around is for this issue with 3x with multiple repairs. ???
Marshall
On 2017-10-07 10:41, "Steinmaurer, Thomas" <th...@dynatrace.com> wrote:
> Marshall,
>
> -pr should not be used with incremental repairs, which is the default since 2.2. But even when used with full repairs (-full option), this will cause troubles when running nodetool repair -pr from several nodes concurrently. So, unfortunately, this does not seem to work anymore and makes the operational side much more difficult to handle.
>
> We are in the same boat. There has been a lengthy discussion on that a few weeks ago and was asked to open the following ticket:
> https://issues.apache.org/jira/browse/CASSANDRA-13885
>
> Thomas
>
> -----Original Message-----
> From: marshall.s.king@virtustream.com [mailto:marshall.s.king@virtustream.com]
> Sent: Samstag, 07. Oktober 2017 19:28
> To: user@cassandra.apache.org
> Subject: Got error, removing parent repair session - When doing multiple repair -pr ââ¬â Cassandra 3.x
>
>
>
> On cassandra 3.11 were getting an error (see error below) when doing multiple repairs with -pr option. Performing this type of repair worked fine on Cassandra 2.x. My questions are:
>
> Is this a valid operation (nodetool repair -pr .. simultaneously on multiple nodes) to do?
> If question one is valid, then has anyone seen this error before and if so what do you do to fix?
>
> The error is:
>
> ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 RepairMessageVerbHandler.java:168 - Got error, removing parent repair session ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 CassandraDaemon.java:228 - Exception in thread Thread[AntiEntropyStage:1,5,main] java.lang.RuntimeException: Parent repair session with id = 2689fa80-a2d5-11e7-9a55-b1319b7b5400 has failed.
>
> Thanks, Marshall
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
> For additional commands, e-mail: user-help@cassandra.apache.org
>
> The contents of this e-mail are intended for the named addressee only. It contains information that may be confidential. Unless you are the named addressee or an authorized designee, you may not copy or use it, or disclose it to anyone else. If you received it in error please notify us immediately and then destroy it. Dynatrace Austria GmbH (registration number FN 91482h) is a company registered in Linz whose registered office is at 4040 Linz, Austria, FreistädterstraÃe 313
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
> For additional commands, e-mail: user-help@cassandra.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
For additional commands, e-mail: user-help@cassandra.apache.org
RE: Got error, removing parent repair session - When doing multiple repair -pr — Cassandra 3.x
Posted by "Steinmaurer, Thomas" <th...@dynatrace.com>.
Marshall,
-pr should not be used with incremental repairs, which is the default since 2.2. But even when used with full repairs (-full option), this will cause troubles when running nodetool repair -pr from several nodes concurrently. So, unfortunately, this does not seem to work anymore and makes the operational side much more difficult to handle.
We are in the same boat. There has been a lengthy discussion on that a few weeks ago and was asked to open the following ticket:
https://issues.apache.org/jira/browse/CASSANDRA-13885
Thomas
-----Original Message-----
From: marshall.s.king@virtustream.com [mailto:marshall.s.king@virtustream.com]
Sent: Samstag, 07. Oktober 2017 19:28
To: user@cassandra.apache.org
Subject: Got error, removing parent repair session - When doing multiple repair -pr — Cassandra 3.x
On cassandra 3.11 were getting an error (see error below) when doing multiple repairs with -pr option. Performing this type of repair worked fine on Cassandra 2.x. My questions are:
Is this a valid operation (nodetool repair -pr .. simultaneously on multiple nodes) to do?
If question one is valid, then has anyone seen this error before and if so what do you do to fix?
The error is:
ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 RepairMessageVerbHandler.java:168 - Got error, removing parent repair session ERROR [AntiEntropyStage:1] 2017-09-26 16:10:41,135 CassandraDaemon.java:228 - Exception in thread Thread[AntiEntropyStage:1,5,main] java.lang.RuntimeException: Parent repair session with id = 2689fa80-a2d5-11e7-9a55-b1319b7b5400 has failed.
Thanks, Marshall
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
For additional commands, e-mail: user-help@cassandra.apache.org
The contents of this e-mail are intended for the named addressee only. It contains information that may be confidential. Unless you are the named addressee or an authorized designee, you may not copy or use it, or disclose it to anyone else. If you received it in error please notify us immediately and then destroy it. Dynatrace Austria GmbH (registration number FN 91482h) is a company registered in Linz whose registered office is at 4040 Linz, Austria, Freistädterstraße 313
---------------------------------------------------------------------
To unsubscribe, e-mail: user-unsubscribe@cassandra.apache.org
For additional commands, e-mail: user-help@cassandra.apache.org
Re: Got error, removing parent repair session - When doing multiple repair -pr — Cassandra 3.x
Posted by kurt greaves <ku...@instaclustr.com>.
Nvm forgot about that
<https://issues.apache.org/jira/browse/CASSANDRA-13885> ticket and that
full repairs do anti-compactions now. For anyone else who doesn't already
know, what I said above is true but full repairs won't help the multiple
concurrent repair issue because they do anti-compactions as well now...
On 9 October 2017 at 01:42, kurt greaves <ku...@instaclustr.com> wrote:
> Note 3.11 uses incremental repair by default. If you were using full
> repairs previously you'll need to specify -full as well when repairing in
> 3.11.
>
> Incremental repair won't work if you run it on multiple nodes at the same
> time that have replica overlap in the token ring.
>
>
Re: Got error, removing parent repair session - When doing multiple repair -pr — Cassandra 3.x
Posted by kurt greaves <ku...@instaclustr.com>.
Note 3.11 uses incremental repair by default. If you were using full
repairs previously you'll need to specify -full as well when repairing in
3.11.
Incremental repair won't work if you run it on multiple nodes at the same
time that have replica overlap in the token ring.