Hi Ricardo,

It's working. Thanks for the quick fix! :)

Now I could boot my first SGE cluster in switchengines. I only had a weird
issue when booting it because ansible crashed restarting iptables. It
crashed in this handler:
https://github.com/gc3-uzh-ch/elasticluster/blob/master/elasticluster/share/playbooks/roles/iptables/handlers/main.yml#L3

here you have the full ansible error output:

RUNNING HANDLER [iptables : reload iptables]
***********************************
fatal: [compute001]: FAILED! => {"failed": true, "msg": "the field 'args'
has an invalid value, which appears to include a variable that is undefi
ned. The error was: 'ansible.parsing.yaml.objects.AnsibleUnicode object'
has no attribute u'Debian'\n\nThe error appears to have been in '/opt/ela
sticluster/src/elasticluster/share/playbooks/roles/iptables/handlers/main.yml':
line 3, column 3, but may\nbe elsewhere in the file depending on t
he exact syntax problem.\n\nThe offending line appears to be:\n\n\n- name:
reload iptables\n  ^ here\n"}
fatal: [compute002]: FAILED! => {"failed": true, "msg": "the field 'args'
has an invalid value, which appears to include a variable that is
undefined. The error was: 'ansible.parsing.yaml.objects.AnsibleUnicode
object' has no attribute u'Debian'\n\nThe error appears to have been in
'/opt/elasticluster/src/elasticluster/share/playbooks/roles/iptables/handlers/main.yml':
line 3, column 3, but may\nbe elsewhere in the file depending on $he exact
syntax problem.\n\nThe offending line appears to be:\n\n\n- name: reload
iptables\n  ^ here\n"}
fatal: [frontend001]: FAILED! => {"failed": true, "msg": "the field 'args'
has an invalid value, which appears to include a variable that is
unde$ined. The error was: 'ansible.parsing.yaml.objects.AnsibleUnicode
object' has no attribute u'Debian'\n\nThe error appears to have been in
'/opt/e$asticluster/src/elasticluster/share/playbooks/roles/iptables/handlers/main.yml':
line 3, column 3, but may\nbe elsewhere in the file depending on the exact
syntax problem.\n\nThe offending line appears to be:\n\n\n- name: reload
iptables\n  ^ here\n"}
        to retry, use: --limit
@/opt/elasticluster/src/elasticluster/share/playbooks/site.retry

but after the crash I retried "elasticluster setup sge" and it worked fine
and I could login and submit my first jobs.

any idea why this could happen?



2016-11-11 14:01 GMT+01:00 Riccardo Murri <[email protected]>:

> On 11 November 2016 at 11:37, Riccardo Murri <[email protected]>
> wrote:
> > Hi Pablo,
> >
> >>> I have tested elasticluster with SWITCHengines (openstack) and I hit a
> >>> problem when booting my first test cluster. The nodes are started but
> I get
> >>> this error when elasticluster checks the ssh connection to nodes:
> >>> https://gist.github.com/pescobar/f75b561f5775f20d60139bca9cbc4a47
> >>> [...]
> >>> My guess is that elasticluster is not parsing correctly the public ip.
> is
> >>> it possible? Did anyone hit this issue?
> >>
> >> I think that your guess is right - This looks certainly a bug in
> >> ElastiCluster, in that it mistakes an IPv6 address for IPv4+port ...
> >
> > Looks like the regexp used to match IPv6 addresses is incorrect::
> >
> >   >>> a1 = '2001:620:5ca1:1f0:f816:3eff:fe05:f40f'
> >   >>> IPV6_RE = re.compile('\[([a-f:A-F0-9]*[%[0-z]+]?)\](?::(\d+))?')
> >   >>> m = IPV6_RE.match(a1)
> >   >>> print(m)
> >   None
> >
> > I'll try to fix that ASAP.
>
> I have just pushed a fix in commit 165e20224d37a1d33fb2bd11218352;
> it works for me but as I said I cannot really test because no IPv6 here.
>
> Can you please try and let me know?
>
> Note that this introduces an additional dependency on Python package
> `netaddr` so you might need to run `pip install -e .` after update the
> repo.
>
> Ciao,
> R
>

-- 
You received this message because you are subscribed to the Google Groups 
"elasticluster" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to