Comment 15 for bug 1327763

Thank you for your answer, and sorry for being so late :(
but we're very occupied... ...maybe you're also familiar with this
situation: more work, less workers...

Well, I think, we already could produce the right output, and also being
in contact with percona, they provided a bugfix, which seems to work.
All in all i think it's a combination of several factors, that we could
reduce the crashes to a very minimum of about less than once a week on
one of the nodes.

What we did is:
1. put the database on SSDs
2. a dedicated switch for synchronization of the nodes
3. doing writes mainly on only one node

I will put this info also on launchpad,
so maybe some other people can benefit from...

Mit freundlichen Grüßen (kind regards)

Bert Hutzler

Quoka GmbH
EDV Abteilung
Chemiestraße 14-15
68623 Lampertheim

Telefon: +49 (0) 6206 / 931-128
Telefax: +49 (0) 6206 / 931-149
E-Mail: mailto:<email address hidden>
Internet: http://www.quoka.de

AG Darmstadt - Reg. Abt. Lampertheim - HRB 61482 - USt-ID-Nr. DE 175 723 521 - Geschäftsführung: Patrick Günter

On 22.03.2015 22:14, dannie wrote:
> I'm back with some new info. After the last update to 5.6.22-72.0-56
> every node crashes within 3-4h, and at random times. After some help
> from the percona crew to get the debug info this is what I got, hope it
> helps.
>
> Let me know if there is anything else I can post
>
>
> ** Attachment added: "crash.log"
> https://bugs.launchpad.net/percona-xtradb-cluster/+bug/1327763/+attachment/4352525/+files/crash.log
>