#theforeman

It looks like foreman-installer installs postgresql server on the machine where you install Foreman - even when specifying a server that is not localhost. Is there an option in foreman-installer to disable the installation of psql server? Can't seem to find anything in the docs / manpage.

dLobatog: no havent touched it since i looked at the wrong DB query

rabbel: --foreman-db-manage=false which will disable both installation and migration/seed of the DB, you're expected to do it

Dominic: 'you're expected to do it' as in I'd have to run all rake migration scripts myself after installation?

yes

Interesting. Ok, I'll leave it as is for now then :-)

dLobatog: if I change the type in the DB it works, but then many other things fail, so I gave up. My thought was to change the query from: [WHERE `hosts`.`type` IN ('Host::Managed')] to: [WHERE `hosts`.`type` IN ('Host::Managed') OR ('Host::Discovered')]

Dominic: any plans to change this behaviour in the future?

rabbel: there's no current PR to change it

Ok, thx!

dLobatog: the current query is looking for a type of Hosts::Managed. But since its a discovered host the type is actually Hosts::Discovered and thats where error happens I believe

rabbel: with foreman-rake commands its relatively easy to migrate+seed into another db (i've tried this with mysql, but on my test box)

devmodem: I see. Well, I'm not trying to move or migrate anything. I just ran foreman-installer with an external psql db specified. I was surprised to see the process running on the instance itself.

rabbel: ah, understood now. btw, with migrate i meant running initial migrations to create tables etc.

Singularity42: yeah I found a few other problems , in any case, a workaround to get the right OS and templates is to create the discovered host with a hostgroup that has them

dLobatog: oh no kidding huh. let me give that a try

dLobatog: hmm. same issue. let me go back and double check the new hostgroup

dLobatog: ya, same issue

Singularity42: if you click on the 'resolve' button, it will say that they are not resolved, but clicking on submit works, they're assigned

dLobatog: which is the same actions as before, but the problem is when the host gets to looking for its unattended script it will fail to find it. Let me get up to that point and make sure the error is the same.

I have a newly installed foreman using an on demand download policy. When trying to install a new system using my product repo I get failures that cause the install to stop. When I restart the install it goes a little bit further next time before stopping..

Is there a way to preload the repo?

dLobatog: they looked similar so I thought they went hand in hand. But maybe you would know better - https://gist.github.com/anonymous/08d83bb95c741...

dLobatog: this only happens on hosts that cant resolve their templates before hitting submit

dLobatog: if I go to the host after clicking submit, as its sitting in Build: Pending Installation, and review the provisioning template, it renders fine.

Dominic: wooo - http://projects.theforeman.org/issues/12111#cha...

Anyone know what'd cause the Type field in Content Hosts to not update? should be type vmware, and CH even coorelated the Vhost it's running on: http://pasteboard.co/8W0bBHl1K.png

anyone provisioning image based vms with libvirt? the vm will say that there's nothing to boot in the hard disk

igorgarciaruiz: I use qcow base images on libvirt, seems fine. you've got a working base image that boots, right?

yes gwmngilfen:, in my case it's a .img, and the only difference I see is that the original vm has a "SATA disk" and the one created from foreman "VirtIO disk". although, not really sure whether that has anything to do

is there a way to associate a location to an activation key in katello?

igorgarciaruiz: if you check the xml for the new vm, does it list the base image as a backing store correctly? should say something like <source file='/path/to/file'/>

although I have a feeling backing stores *have* to be QCOW

just destroyed it :). I'm going to use qcow2 and have a look at the xml

sounds good, let me know ;)

i can dig into my foreman setup to compare notes if you still get stuck

what else can i look into if I have ESX hosts coming back during provisioning that say "http://puppet/unattended/provision;) connection failed - Error 404" I have troubleshooted that error many times in the past, but this one I cant seem to resolve

thanks gwmngilfen:!. Actually, the source file lists itself, not the base image

is that within a <backingstore/> element or above it?

if the latter, you're not using the base image at all, which sounds like a setup issue

I don't see a <backingstore/> element :(. http://termbin.com/w064

yeah, thats not using the base then. In foreman, go to your compute resource, images tab, edit the appropriate image you're trying to deploy - have you set "image path" ?

it should be the full path to the base image on the hypervisoer

eg i have /var/lib/libvirt/images/debian-jessie-7.0.0-3-amd64.qcow2

(yeah, the 7.0 is a typo :P)

along with it, the only thing I see in the production logs is - unable to find provision; template for currahee.deploy.lab.beer.town running ESXi6.0UbAutomation 6.02 - but when i render it in Foreman GUI it works fine

Hi

Singularity42: can you get debug logs for that - disparity between sppof render and real render usually points to an orchestration error

*spoof

Anyone can give me a hints were I can get more info on why my newly created host seem frozen on "Pending installation" step ?

allright. so the vm path seems to be fine, but I see that at this point the compute profile I'm using isn't matching the type of storage that the image is using. so I'll give it another go

ah yeah, that could be it