[slurm-users] Slurm 19.05 X11-forwarding

Sat Feb 29 04:56:26 UTC 2020

Hi all,
First off, thank you all for all of your quick replies and suggestions on how to solve this problem of mine.
With some additional help from Tina  Friedrich  I did a test proposed by her:

First step:  ”ssh -X <node>”.
Second step: ”ssh -X localhost”.
Third step: ”srun —x11 <script_to_be_performed_at_node>”

(It should read a double ” -” but my mail-sw on this device doesnt let me do that, sorry!)
This actually worked and I got the X-application to be displayed at my working x-server/display.

So to summarize, I had to ssh to the node and when connected I ssh’d once more to the localhost (i.e the localhost at the node) and then started a ”srun”-session. This did work, I got the X-forwarding to work and sent the X-application back to my available X-sever/screen.

This isn’t really the way I aspected the srun nor X11-forwarding to work. I thought that I could run the srun-command with X11-forwarding called from an sbatch-jobarray-script and get the X11-forwarding to my display. Yes I know that having 4000 nodes will pose a problem with a lot of X11-forwarding being sent, but that will not be a problem for me with only a handful of nodes and utlizing e.g xvfb.
I’ll work more on this matter now that I know a way of how to get it to work.

Thank you!
Best regards,
Pär Lundö

________________________________
From: "slurm-users" <slurm-users-bounces at lists.schedmd.com>
Sent: 28 feb. 2020 17:52
To: "wagner at itc.rwth-aachen.de" <wagner at itc.rwth-aachen.de>
Cc: "slurm-users at lists.schedmd.com" <slurm-users at lists.schedmd.com>
Subject: Re: [slurm-users] Slurm 19.05 X11-forwarding

Hi Marcus,

You are correct, but it doesnt show anything regarding the X11-forwarding.

Thank you for your input!

Best regards,
Pär Lundö
________________________________
From: "slurm-users" <slurm-users-bounces at lists.schedmd.com>
Sent: 28 feb. 2020 15:57
To: "slurm-users at lists.schedmd.com" <slurm-users at lists.schedmd.com>
Subject: Re: [slurm-users] Slurm 19.05 X11-forwarding

Hi Pär,

yes , you can -v or e.g. -vvv together with srun. I'm not sure, though, if taht shows anything X11-related, but you might try.

Best
Marcus

On 2/28/20 3:45 PM, Pär Lundö wrote:

Hi everyone

Thank you for your support.
I’ve done a few changes and done some further testing but it has not solved my problem.
Regardless of the settings for the sshd I can’t get it to
I am able to use SSH to the node directly along with ”-X”-argument and get the X11 forwarding to work.
Is there a debug output from srun to retrieve during the connection to the node?
(Surely the simple answer to that is yes! I’ll try and locate that debug-function, or can someone help with that?)

Best regards,
Pär Lundö

________________________________
From: "slurm-users" <slurm-users-bounces at lists.schedmd.com><mailto:slurm-users-bounces at lists.schedmd.com>
Sent: 27 feb. 2020 12:29
To: "Slurm User Community List" <slurm-users at lists.schedmd.com><mailto:slurm-users at lists.schedmd.com>
Subject: Re: [slurm-users] Slurm 19.05 X11-forwarding

Hi Sean,

Thank you for your reply.
I will test it asap.

Best regards,
Pär Lundö
From: slurm-users <slurm-users-bounces at lists.schedmd.com><mailto:slurm-users-bounces at lists.schedmd.com> On Behalf Of Sean Crosby
Sent: den 27 februari 2020 10:26
To: Slurm User Community List <slurm-users at lists.schedmd.com><mailto:slurm-users at lists.schedmd.com>
Subject: Re: [slurm-users] Slurm 19.05 X11-forwarding

I remember that we had to add this to our /etc/ssh/sshd_config to get X11 to work with Slurm 19.05

X11UseLocalhost no

We added this to our login nodes (where users ssh to), and then restarted the ssh server. You would then need to log out and log back in with X11 forwarding again.

Sean

--
Sean Crosby | Senior DevOpsHPC Engineer and HPC Team Lead
Research Computing Services | Business Services
The University of Melbourne, Victoria 3010 Australia

On Wed, 26 Feb 2020 at 20:52, Pär Lundö <par.lundo at foi.se<mailto:par.lundo at foi.se>> wrote:
Hi,

Thank you for your quick replies.
Please bear with me as I am a newbie regarding Slurm and Linux.
My hostname is not a FQDN and I´m running Slurm on a local node (slurmctld and slurmuser are the same) just to verify that the X11-forwarding is working (amongst other things).
The output of "xauth list" is as follows:
"
parlun-E6540/unix:  MIT-MAGIC-COOKIE-1  bcf263421d143e6ba3297dfe926c68b8
#ffff#7061726c756e2d4536353430#:  MIT-MAGIC-COOKIE-1  bcf263421d143e6ba3297dfe926c68b8
"
I seem to be missing any ":0", i.e. to wich display or Xserver my user (parlun-E6540) uses?
I´ve tried adding the user "slurm", who is designated slurmuser, via "xhost +SI:localuser:slurm" but I still get the error stating that the magic cookie could not be retrieved.

You are correct, X11Parameters only holds one option "home_xauthority".
Does this setting automatically search in the "/home/slurm"-dir for the Xauthority-file? (Given that the slurmuser is "slurm").
Trying to locate the storage where xauthoriy is present, I do not find it my home-dir instead it is in some other OS-dir.

Searching for additional help I came across a debian package (xvfb) for virtual framebuffers. i.e. video/GUI is sent to a virtual buffer instead of a physical display. Anyone tried that?
I haven´t not yet installed it, but is curious if that´s a way to go.

Best regards,
Pär Lundö

-----Original Message-----
From: slurm-users <slurm-users-bounces at lists.schedmd.com<mailto:slurm-users-bounces at lists.schedmd.com>> On Behalf Of Ryan Novosielski
Sent: den 25 februari 2020 21:23
To: slurm-users at lists.schedmd.com<mailto:slurm-users at lists.schedmd.com>
Subject: Re: [slurm-users] Slurm 19.05 X11-forwarding

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I seem to remember there being a config option to specify rewriting the hostname as well. I thought it was part of X11Parameters, but I only see one option there:

https://slurm.schedmd.com/archive/slurm-19.05-latest/slurm.conf.html

On 2/25/20 10:55 AM, Tina Friedrich wrote:
> I remember having issues when I set up X forwarding that had to do
> with how the host names were set on the nodes. I had them set (CentOS
> default) to the fully qualified hostname, and that didn't work - with
> an error message very similar to what you're getting, if memory serves
> right. 'Fixed' it by setting the hostnames on all my cluster nodes to
> the short version.
>
> What does 'xauth list' give you on your nodes (and the machine you're
> coming from)?
>
> This is/was SLURM 18.08 though, not sure if that makes a difference.
>
> Tina
>
> On 25/02/2020 04:55, Pär Lundö wrote:
>> Hi,
>>
>> Thank you for your reply Patrick. I´ve tried that but I still get the
>> error stating that the magic cookie could not be retrieved.
>> Reading Tim´s answer, this bug should have been fixed in a release
>> following 18.08, but I´m using 19.05 thus it should have been taken
>> care of?
>>
>> Best regards, Pär Lundö
>>
>> -----Original Message----- From: slurm-users
>> <slurm-users-bounces at lists.schedmd.com<mailto:slurm-users-bounces at lists.schedmd.com>> On Behalf Of Patrick Goetz
>> Sent: den 24 februari 2020 21:38 To:
>> slurm-users at lists.schedmd.com<mailto:slurm-users at lists.schedmd.com> Subject: Re: [slurm-users] Slurm
>> 19.05 X11-forwarding
>>
>> This bug report appears to address the issue you're seeing:
>>
>> https://bugs.schedmd.com/show_bug.cgi?id=5868
>>
>>
>>
>> On 2/24/20 4:46 AM, Pär Lundö wrote:
>>> Dear all,
>>>
>>> I started testing and evaluating Slurm roughly a year ago and used
>>> it succesfully with MPI-programs. I have now identified that I need
>>> to use X-forwarding in order to make use of an application needed to
>>> run a GUI. However I seem not to be able to use the "X11Forwarding"
>>> in Slurm.conf with any sucess. A simple test with srun as "srun -
>>> -x11 <script-to-start>", yields an error stating: "srun: error:
>>> x11_get_xauth: Could not retrieve magic cookie. Cannot use X11
>>> forwarding." I'm guessing that there some additional setting I'm
>>> missing out on.
>>> Searching through documentation does not reveal anything
>>> extraordinary to be performed with the X11-forwarding. I have
>>> searched bug-reports and found similiar problems but fixes have been
>>> implemented in Slurm-versions prior to the one I am using.
>>> In addition users have been reporting that the implementation made
>>> by Slurm, now have fixed their issues. I have used the
>>> "export=DISPLAY, HOME" as an additional argument for srun but
>>> without any progress. Anyone with similiar problem who can aid or
>>> advice me on howto use the X11Forward feature? Any help is much
>>> appreciated.
>>>
>>> I am running Slurm 19.05 and Ubuntu 18.10.
>>>
>>> Best regards, Pär Lundö
>>>
>>> This message is from an external sender. Learn more about why this
>>> matters.
>>> <https://ut.service-now.com/sp?id=kb_article&number=KB0011401>
>>>
>>>
>>

- --
 ____
 || \\UTGERS,     |----------------------*O*------------------------
 ||_// the State  |    Ryan Novosielski - novosirj at rutgers.edu<mailto:novosirj at rutgers.edu>
 || \\ University | Sr. Technologist - 973/972.0922 ~*~ RBHS Campus
 ||  \\    of NJ  | Office of Advanced Res. Comp. - MSB C630, Newark
      `'
-----BEGIN PGP SIGNATURE-----

iF0EARECAB0WIQST3OUUqPn4dxGCSm6Zv6Bp0RyxvgUCXlWCNgAKCRCZv6Bp0Ryx
vjTSAJ9S7WA/rYjw5ylOljq9IpPqwTsHKACfbn/nnGJvE+QHvJoJ2djkkCQjMok=
=Ju41
-----END PGP SIGNATURE-----

--
Marcus Wagner, Dipl.-Inf.

IT Center
Abteilung: Systeme und Betrieb
RWTH Aachen University
Seffenter Weg 23
52074 Aachen
Tel: +49 241 80-24383<tel:+49%20241%2080-24383>
Fax: +49 241 80-624383<tel:+49%20241%2080-624383>
wagner at itc.rwth-aachen.de<mailto:wagner at itc.rwth-aachen.de>
www.itc.rwth-aachen.de<http://www.itc.rwth-aachen.de>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.schedmd.com/pipermail/slurm-users/attachments/20200229/6003944d/attachment-0001.htm>