Open sebastian-luna-valero opened 1 year ago
Hi, do you mean it should be invalid because of the following?
external comes without HTCondor now, this is because the configuration of HTCondor changed and is best done with the getHTCondor-Script. With that script you can set your own cluster password and host. This would require a lot of error-prone configuration otherwise.
I cannot answer immediately whether it is still valid or not. In addition, I do not know what
require a lot of error-prone configuration otherwise.
precisely means. @mira-miracoli (currently not available) could easily answer your second question I believe. With more time I could have a look into the first.
Yes, that's what I mean. Let's then wait for Mira's thoughts on this.
Thank you!
@mtangaro could you please confirm whether you used this Ansible role to deploy HTCondor 10.x?
At EU, we are not using HTCondor 10. x. We are still using the older version 8.8.15
.
@mtangaro could you please confirm whether you used this Ansible role to deploy HTCondor 10.x?
I've deployed a Pulsar endpoint, which is working fine, using:
condor_q --version
$CondorVersion: 10.5.0 2023-06-05 BuildID: 650732 PackageID: 10.5.0-1 $
$CondorPlatform: x86_64_AlmaLinux9
Thanks, All.
Then, apparently I misinterpreted the comments in https://github.com/usegalaxy-eu/vgcn/pull/56 and assumed this role wasn't properly maintained.
In that sense, I invested some time updating a separate Ansible role for HTCondor in https://github.com/grycap/ansible-role-htcondor/pull/10. Sorry for the duplication of effort. I am planning to use this updated Ansible role for HTCondor streamline the deployment of pulsar via https://im.egi.eu/ in the context of the EuroScienceGateway project.
Oh this other role was updated the last time 3 years ago. I hope they merge it. Let us know if you want to use this role here and need anything else.
Just for the record, check also https://github.com/usegalaxy-eu/vgcn/commit/ebec04289b3d08c1d7ecd8b11f8c549dfcf4e58b
@bgruening @sanjaysrikakulam Actually why don't we switch to grycap/ansible-role-htcondor also at usegalaxy.eu? I have been doing some work on our role before realizing the other role exists, but honestly this other role is still better than ours + new changes.
If we are happy with that other role I suggest we even archive this repository.
@kysrpex it was hardly maintained for 3 years. If its back to live and we can use it, lets do it.
@kysrpex it was hardly maintained for 3 years. If its back to live and we can use it, lets do it.
That role seems to do things that ours + new changes does not, most importantly installing two roles on the same machine (needed for the headnode) and supporting the "release" and "stable" channels. If I commit what I have it will be a half-assed solution compared to said role. I am giving it a try, if it does not work we can always continue using ours.
@kysrpex it was hardly maintained for 3 years. If its back to live and we can use it, lets do it.
That role seems to do things that ours + new changes does not, most importantly installing two roles on the same machine (needed for the headnode) and supporting the "release" and "stable" channels. If I commit what I have it will be a half-assed solution compared to said role. I am giving it a try, if it does not work we can always continue using ours.
Just FYI, after looking at the role I am not so sure after all. They're quite different approaches... I'll update you with whatever I do anyway.
We are using now kysrpex/ansible-htcondor-grycap (but should be using usegalaxy-eu/ansible-htcondor-grycap, see usegalaxy-eu/infrastructure-playbook#1234).
Considering that, I would suggest archiving this repository.
Hello,
After https://github.com/usegalaxy-eu/vgcn/pull/56 could you confirm this Ansible role is no longer valid?
Is it worth investing effort in updating it to streamline pulsar deployments?
Best regards, Sebastian