#theforeman

using the foreman API how can I get a configuration related items ie: ssl cert full path to .pem etc?

hello i have a question about sc-params ... how can i list with hammer the value of a sc-param for a host?

aka override value

Im faced with a lot of Katello Tasks in state paused and repo synchronization has failed during the last 10 days. I do not quite know where to start debugging, has anybody some ideas?

rauchrob: the tasks/ page would be a good point, which tasks are failing?

dLobatog: Lots of, Around 800 in status paused, starting from June 13.

dLobatog: Most of of them are "Synchronize Repository", "Generate Capsule Metadata and Sync" and "Metadata generate"

dLobatog: But I also see that most of these tasks (around 750) occured just when our repo sync problems started (about 10 days ago)

rauchrob: what are the errors you see? going to the dynflow console might help you http://www.katello.org/troubleshooting/

dLobatog: the "Synchronize Repository" tasks, which started to fail around 10 days ago all have the error message "Trying to use logger katello/pulp_rest which has not been configured.

dLobatog: I have applied a workaround for this yesterday: https://groups.google.com/forum/#!topic/foreman...

rauchrob: what katello vesion do you have

?

inecas: 2.4.1

rauchrob: here is the issue http://projects.theforeman.org/issues/15114/

After applying the patch from https://groups.google.com/forum/#!topic/foreman..., I can resume the "Synchronize Repository" tasks without the error occuring again, but they are kind of stuck at 71%

rauchrob: it has been fixed in 3.0.2, this happens at the foremant-tasks startup

rauchrob: you added :pulp_rest: :enabled: true to katello.yaml right?

rauchrob: for the tasks, that are in paused state due to this, resuming it should help

dLobatog: right

inecas: do you mean I should apply some patch and resume the tasks?

rauchrob: just reusuming the tasks should be ok

rauchrob: the issue really is when the foreman-tasks service is restarting

rauchrob: which leads me to another potential cause…

inecas: as I said, after resuming they are stuck at 71%

rauchrob: always?

yes

I also noticed some problems with the pulp-workers yesterday (they were not running). I was able to start them manually with systemctl start pulp_woker-{0,1}

rauchrob: what's the output of hammer ping?

dLobatog: Everything ok

:)

rauchrob: I think the pulp_rest logger might be just a side-issue, but not the real reason for the stuck tasks all the time

inecas: yes...

rauchrob: could you share tail (let's say last 1000 lines) of /var/log/foreman/dynflow_executor.output log?, just to see if there is not some unexpected restarting there

inecas: yes, just a second

inecas: http://pastebin.com/uE2vaefM

rauchrob: ok, that messages don't indicate something unexpected (that happne usually when the foreman-tasks service is not running, which might happne on services or servers restart

So, these 'No executor available' errors are normal?

ineacs: is it possible to somehow just cancel all these paused jobs and start fresh?

rauchrob: if the tasks sevice is not running, bug once the tasks service gets started, it pickes the tasks up

rauchrob: it's recommand to resume the paused tasks

inecas: so.. then should i resume just everything?

rauchrob: as the task stuck in the middle could lead to some data inconsisten state

rauchrob: there should be hammer command for that

rauchrob: sry, need to go afk for a while

inecas: so you would suggest resuming all of them?

inecas: ... by using hammer

dLobatog: Well, after running 'hammer task resume', of my 800 paused/error tasks only 36 remained (only very "old ones")

happy sysadmin day everyone :)

gwmngilfen: thank you. you too

my last day before summer hollyday

what is the status of ovirt 4.0 integration. was is fixed with 1.12.1?

no, the ticket is still open/new #15560

no: #15560 is http://theforeman.org/issues/15560 "Bug #15560: oVirt 4.0 API is not supported - Foreman"

thx Dominic