On 01/13/2016 06:20 AM, Raghavendra Talur wrote:


On Wed, Jan 13, 2016 at 3:49 PM, Niels de Vos <[email protected]
<mailto:[email protected]>> wrote:

    On Wed, Jan 13, 2016 at 10:35:42AM +0100, Xavier Hernandez wrote:
    > The same has happened toslave34.cloud.gluster.org 
<http://slave34.cloud.gluster.org>. I've
    disabled it to
    > allow regressions to be run on other slaves.
    >
    > There are two files owned by root inside
    > /home/jenkins/root/workspace/rackspace-regression-2GB-triggered:
    >
    > -rwxr-xr-x  1 root    root     10124 Jan  7 17:54 file_lock
    > drwxr-xr-x  3 root    root      4096 Jan  7 18:31slave34.cloud.gluster.org 
<http://slave34.cloud.gluster.org>:

    Thanks!

    I've looked into this a little more now, and might have identified the
    problem.

    This one failed with an unrelated error:

    
https://build.gluster.org/job/rackspace-regression-2GB-triggered/17413/console

       ...
       Building remotely on slave34.cloud.gluster.org
    <http://slave34.cloud.gluster.org> (rackspace_regression_2gb) in
    workspace
    /home/jenkins/root/workspace/rackspace-regression-2GB-triggered
        > git rev-parse --is-inside-work-tree # timeout=10
       Fetching changes from the remote Git repository
        > git config remote.origin.url
    git://review.gluster.org/glusterfs.git
    <http://review.gluster.org/glusterfs.git> # timeout=10
       Fetching upstream changes from
    git://review.gluster.org/glusterfs.git
    <http://review.gluster.org/glusterfs.git>
       ...

    The next run on slave34 failed because of the weird directory:

    
https://build.gluster.org/job/rackspace-regression-2GB-triggered/17440/console

       ...
       Building remotely on slave34.cloud.gluster.org
    <http://slave34.cloud.gluster.org> (rackspace_regression_2gb) in
    workspace
    /home/jenkins/root/workspace/rackspace-regression-2GB-triggered
       Wiping out workspace first.
       java.io.IOException: remote file operation failed:
    /home/jenkins/root/workspace/rackspace-regression-2GB-triggered at
    hudson.remoting.Channel@62ecdacb:slave34.cloud.gluster.org
    <http://slave34.cloud.gluster.org>:
       ...

    Note the "Wiping out workspace first." line. This comes from an option
    in the regression job. This seems to be a recently added "Additional
    Behaviour" in the Jenkins job. Did anyone add this on purpose, or was
    that automatically done with a Jenkins update or something?


The three additional behaviour added in regression configuration were
added by me and I simply copied
whatever was there in smoke configuration page. We can try removing this
configuration line but the tests weren't getting started without this(we
got it running after a restart so this might be false requirement).

Technically, it is not a harmful configuration and wiki pages recommend
it. It says such errors occur only if files were created and left
open/locked by some tests or were created with different permissions. We
still need to identify what tests are responsible for this.


I have removed this option as regression tests were continuously failing. Please feel free to re-enable as deemed necessary.

Thanks,
Vijay

_______________________________________________
Gluster-infra mailing list
[email protected]
http://www.gluster.org/mailman/listinfo/gluster-infra

Reply via email to