perfSONAR User Q&A and Other Discussion

[Andrew Lake <>]

Hi,

Hmmm the psconfig client doesn’t have too many changes in 5.0. Does “psconfig validate http://perf-pwa.sanren.ac.za/pub/config/sanren_mesh_2019.json” work? Is there anything interesting in /var/log/perfsonar/psconfig-pscheduler-agent.log or /var/log/perfsonar/psconfig-pscheduler-agent-transactions.log?

Thanks,

Andy

On April 19, 2023 at 3:42:00 AM, Johann Hugo () wrote:

Hi all

I'm using the default toolkit distro. Besides fixing the limits file, is there anything else that needs to be done? 

It looks like psconfig is not fetching its mesh configs from my PWA server any more.

[root@ps-16-cpt-is ~]# psconfig remote list
=== pScheduler Agent ===
[
   {
      "url" : "http://perf-pwa.sanren.ac.za/pub/config/sanren_mesh_2019.json",
      "configure-archives" : true
   }
]

psconfig translate can download + display the mesh config file:

[root@ps-16-cpt-is ~]# psconfig translate http://perf-pwa.sanren.ac.za/pub/config/sanren_mesh_2019.json

Zero tasks to run:

[root@ps-16-cpt-is ~]# psconfig pscheduler-stats
Agent Last Run Start Time: 2023/04/19 08:54:00
Agent Last Run End Time: 2023/04/19 08:54:02
Agent Last Run Process ID (PID): 3197
Agent Last Run Log GUID: F6746FEC-DE7E-11ED-A92F-5281C5A1B65D
Total tasks managed by agent: 0

This looks fine to me:
[root@ps-16-cpt-is ~]# pscheduler troubleshoot
Performing basic troubleshooting of ps-16-cpt-is.

ps-16-cpt-is:
  Measuring MTU... 65535 (Local)
  Looking for pScheduler... OK.
  Fetching API level... 5
  Checking clock... Unsynchronized (Not considered fatal)
  Exercising API... Archivers... Contexts... Tests... Tools... OK.
  Fetching service status... OK.
  Checking services... Ticker... Scheduler... Runner... Archiver... OK.
  Checking limits... OK.
  Idle test.... 9 seconds... Finished... Checking archiving... OK.

Regards

Johann

On Tue, Apr 18, 2023 at 5:23 PM Mark Feit <> wrote:

Onno Zweers writes:

 

Today our monitoring warned me that /tmp on our perfsonar node was full. After investigation, I discovered the upgraded packages. I had missed the announcement, apologies for that (too busy to read emails haha).

 

It looks like there are  two unrelated problems:

 

So now we have a system that says “Error loading test listing; measurement archive unreachable:

https://perfsonar-bandwidth.grid.surfsara.nl/esmond/perfsonar/archive/”

                …

I also noptice that opensearch was installed together with perfsonar 5.0. But it failes to start.

 

The error message is a side effect of  OpenSearch having failed to start.  That may have been caused by the full /tmp directory, so I’d recommend making sure /tmp has been cleaned out and trying a “systemctl restart opensearch” to get it going again.  If that fails, please run “journalctl -xe” and send any OpenSearch-related entries here or privately if you’d prefer that.

 

Limit processor is not initialized: Limit 'idleex-default' has unsupported type 'test'

 …

It says “replace” but I can’t seem to find which file or config I need to replace.

 

The limit configuration lives in /etc/pscheduler/limits.conf.  I’ve updated the helper script so it mentions what file it’s checking.

 

--Mark

 

 

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

--

SANReN Engineer

South African National Research Network (SANReN)

National Integrated Cyber Infrastructure System (NICIS)

CSIR NextGen Enterprises and Institutions Cluster

Office: 012 841 2066Email: , Website: www.sanren.ac.za / www.csir.co.za

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