Thor shutting down

Fri May 29, 2015 8:00 pm

Hi

We seem to be having some issues with our Thor cluster. The system seems to be shutting down periodically. I have attached our thormaster Log. If somebody can help point me in the right direction it would be a great help.

Kind regards

David

Fri May 29, 2015 9:25 pm

Hello

Can someone tell us what might be causing the segfault? Slavelog attached.

Also, how do we replace a node with a new one?

Are there instructions?

Kind regards

David

Mon Jun 01, 2015 1:51 pm

David,

Our Operations guys took a look at your logs and told me: "both logs point to 10.12.0.24 having problems...I would try and reboot that node ..if that does not correct the issue re-install the software"

And replacing a node is covered in our Systems Administrators Guide (PDF downloadable here: http://hpccsystems.com/download/docs/installation-and-administration). I found it on page 90.

HTH,

Richard