[A1n_d2n] [EXTERNAL] Singularity path issue impacting analysis -- Reminder of /site, /apps outage on Tuesday

Mon Nov 16 11:00:14 EST 2020

Hi Mingyu and everyone,

It looks that the issues with the removed symlink were resolved, and everything should work again. For future deployments I will make sure to integrate with the module system to (hopefully) avoid this from happening in the future.

Best,
Sylvester

On Nov 15, 2020, at 9:03 PM, Brad Sawatzky <brads at jlab.org<mailto:brads at jlab.org>> wrote:

Hi Mingyu, et al,

It looks like there were some backwards-incompatible changes to
/apps/singularity/ on Friday.  I think the relevant software is still
there, but symlinks need to be made to allow existing code to work
with the new configuration.

I'm CC'ing the group and Sylvester so we all have a heads-up.  I emailed
the IT person in charge and asked that symlinks be restored for now
(along with a reminder that notice of regressions should be given via
scicomp-briefs in the future).  I expect that symlinks will be restored
quickly.  I recommend that folks just wait until tomorrow rather than
make local fixes.

FWIW, it looks like the new standard path will be
 /apps/singularity/3.5.2
instead of
 /apps/singularity/singularity-3.5.2
It looks like IT is also moving towards using the 'module' system.  That
is a good thing in the long run.

Sylvester may want to make the change to invoke the relevant IT module
when he can, or wait a bit and see how this all settles out.

REMINDER:  I'll also take this opportunity to remind everyone of the
 /apps and /site planned outage scheduled that starts Tuesday
 afternoon.  This was posted to the scicomp-briefs list last week.
 More details are here:
   https://mailman.jlab.org/pipermail/jlab-scicomp-briefs/2020q4/000824.html
 Hopefully there is nothing in particular you need to do, but do be
 aware that may be a significant service disruption that evening.

 Note that I will be disabling the /site and /apps mountpoints on
 the Hall C Cluster machines Monday night and restoring them on Wed
 morning.  (Hopefully no one is analyzing on the Hall C machines
 anymore though.)

-- Brad

On Sun, 15 Nov 2020, mingyu at jlab.org<mailto:mingyu at jlab.org> wrote:

Hi Brad,
I encountered a strange problem with hallc_replay. Up to yestoday
(11/14/2020), both hallc replay and root works fine on JLab
ifarm. However, today after I load the PolHe3 software environment:
$ source /group/c-polhe3/software/setup.sh
when I do:
$ root -l
I get this message:
$ /group/c-polhe3/software/opt/hallac_container_1.8/bin/root: line 19:
/apps/singularity/singularity-3.5.2/bin/singularity: No such file or
directory
I am not sure what cause the missing file to prevent root run properly
after setting up the PolHe3 software environment, could you take a look?
Attached screen shot has the details of this problem.
Sincerely,
Mingyu Chen

Hi folks,

I meant to send this out earlier -- sorry for the delay.
 https://hallcweb.jlab.org/wiki/index.php/Pol3He_Analyzer_Git_HowTo

Murchhana kindly beta-tested the instructions for me, so I hope they
work for everyone.  Do let me know if you hit problems or have
questions.

There are many ways to work with git.  The basic idea here is basically
the same as the github 'fork and pull request' workflow.  That is
outlined on the link above.  You can find more detailed descriptions at
the 'Git Howto' page at the bottom of the above link, and/or skim
this quick article.
 https://urldefense.proofpoint.com/v2/url?u=https-3A__reflectoring.io_github-2Dfork-2Dand-2Dpull_&d=DwIBAg&c=CJqEzB1piLOyyvZjb8YUQw&r=BiQgYHr0DbfJNsNM2SdXVw&m=WwDXAOZOy-7KJ4LXBpGJ9JYIE4pWk1qs6GuiLAYLS40&s=TbHBqpzhWF9h1-_mRXkfUe6_9FtTLrGf-JVvWePorxg&e=

(Note that we are using the ANL GitLab server and /not/ GitHub for our
work.  Don't worry though, the process is basically the same.)

-- Brad

--
Brad Sawatzky, PhD <brads at jlab.org<mailto:brads at jlab.org>>  -<>-  Jefferson Lab / Hall C / C111
Ph: 757-269-5947  -<>-  Fax: 757-269-5235  -<>- Pager: brads-page at jlab.org<mailto:brads-page at jlab.org>
The most exciting phrase to hear in science, the one that heralds new
 discoveries, is not "Eureka!" but "That's funny..."   -- Isaac Asimov

_______________________________________________
A1n_d2n mailing list
A1n_d2n at jlab.org<mailto:A1n_d2n at jlab.org>
https://mailman.jlab.org/mailman/listinfo/a1n_d2n

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/a1n_d2n/attachments/20201116/53e965ce/attachment-0001.html>