perfSONAR User Q&A and Other Discussion

[Andrew Lake <>]

Hi,

Sorry for the slow reply. It is not likely an issue with pscheduler or anything like that and probably isolated to the web interface. In addition to /etc/perfsonar/psconfig/pscheduler.d/toolkit-webui.json you should also try deleting /var/lib/perfsonar/toolkit/gui-tasks.conf. It is possible that file somehow got bad data. A couple places to look if that does not work:

- /var/log/perfsonar/web_admin/web_admin.log is where the UI logs errors and such

- The _javascript_ debug console in your browser may have info as well

Also, if this is someplace at LBL I can get a login I’d be happy to take a look as well if nothing above bears fruit.

Thanks,

Andy

On December 1, 2020 at 11:07:18 PM, Rune Stromsness () wrote:

Alternatively, does anyone know where there is documentation on how to re-initialize whatever database has the test configuration?  The web page is completely non-responsive for me and my attempts to clear things out so far:

* Deleting  /etc/perfsonar/psconfig/pscheduler.d/toolkit-webui.json

* Running "pscheduler cancel https://localhost/pscheduler/tasks/" for each task ID

both don't keep all of the jobs from running after a reboot, and don't make the web configuration page usable again.

I suppose I could un-install all of the perfSONAR RPM packages and wipe any perfSONAR directories I can find, but I'm hoping to find something more targeted and/or documented to reset the config.

On 29-Nov-20 21:54, Rune Stromsness (via perfsonar-user Mailing List) wrote:

A fresh new perfSONAR 4.3.2 install on CentOS 7 from this morning was
working as expected for a bit as I manually added new scheduled tests
through the web GUI interface.  Then it has gotten into a state where
any time that I visit the GUI page for configuring scheduled tests it
just spins apparently indefinitely with the Loading notice without any
of the existing tests ever showing up there.  (This is true from various
browsers and browser sessions.)

The troubleshooting information I've found in a few hours of poking
around the web pages and looking through mailing list archives all look
healthy to me, including:

# pscheduler troubleshoot
Performing basic troubleshooting of localhost.

localhost:

  Measuring MTU... 65535 (Local)
  Looking for pScheduler... OK.
  Fetching API level... 4
  Checking clock... OK.
  Exercising API... Status... Tests... Tools... OK.
  Fetching service status... OK.
  Checking services... ticker... scheduler... runner... archiver... OK.
  Idle test.... 8 seconds.... Checking archiving... OK.

pScheduler appears to be functioning normally.

# psconfig pscheduler-stats
Agent Last Run Start Time: 2020/11/29 21:12:42
Agent Last Run End Time: 2020/11/29 21:19:39
Agent Last Run Process ID (PID): 2388
Agent Last Run Log GUID: AC7D646E-32CA-11EB-BB8C-256A68DF5CAF
Total tasks managed by agent: 106
From include files: 106
    /etc/perfsonar/psconfig/pscheduler.d/toolkit-webui.json: 106

psconfig pscheduler-tasks

appears to succeed and show me everything that I had configured.


Restarting perfsonar processes and then rebooting the server haven't
changed the behavior, and I don't see any obvious errors in any logs I
can find.


Any suggestions on how to figure out what is causing this to hang, or
how to recover from this condition?

--
To unsubscribe from this list: https://lists.internet2.edu/sympa/signoff/perfsonar-user

--
To unsubscribe from this list: https://lists.internet2.edu/sympa/signoff/perfsonar-user