Open Stack

On 8/13/15, 6:13 AM, "Jeremy Stanley" <fungi at yuggoth.org> wrote:

>On 2015-08-13 19:38:07 +0800 (+0800), Kai Qiang Wu wrote:
>> I did talked to infra, which I think it is resource issue, But
>> they thought it is nova issue,
>[...]
>
>No, I said the error was being raised by Nova, so was not an error
>coming _from_ the infrastructure we manage. If your jobs are more
>resource-intensive than a typical devstack/tempest job, you'll want
>to look at ways to scale them back.
>
>> It is 20GB disk space, so failed for that.
>
>Correct, we run jobs on resources donated by public service
>providers. Some of them only provide a 20GB root disk. There's
>generally an ephemeral disk mounted at /opt with additional space if
>you can modify your job to leverage that for whatever is running out
>of space.

How large is /opt?

>
>> I think it is related with this, the jenkins allocated VM disk
>> space is not large. I am curious why it failed so often recently.
>> Does os-infra changed something ?
>
>Nothing has been intentionally changed with our disk space on job
>workers as far as I'm aware. Different workers have varying root
>disk sizes depending on the provider where they were booted, but
>they could be as small as 20GB so your job will need to take that
>into account.

20gb isn¹t enough for Magnum¹s CI jobs.  We could link /var/lib/docker to
/opt if there is sufficient space there.

Regards,
-steve

>-- 
>Jeremy Stanley
>
>__________________________________________________________________________
>OpenStack Development Mailing List (not for usage questions)
>Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
>http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev