perfSONAR User Q&A and Other Discussion

[Mark Feit <>]

Brian Tehan writes:

 

I saw that archiving tasks has changed, I ran the  “/usr/lib/perfsonar/archive/perfsonar-scripts/psconfig_archive.sh -n localhost > local_archive.json” command.  I then attempted to run another pscheduler throughput task following the example given but it errors out:

pscheduler task --archive @local_archive.json --slip PT1H --sliprand --repeat-cron "36 0-5 * * *" throughput --source 162.127.1.81 --dest 162.127.1.1 --duration PT30S --bandwidth 3221225472 --ip-version 4 --omit PT5S

Unable to read archive file: [Errno 2] No such file or directory: 'local_archive.json'

 

The bit of code that interprets the “@local_archive.json” strips off the @, does user expansion on the rest (see https://docs.python.org/3/library/os.path.html#os.path.expanduser) and opens whatever path comes out of that.  The text in the error messages says it was correctly interpreted as “local_archive.json” and, while I hate to ask such an obvious question, I need to.  Does that file actually exist?

I am able to setup throughput tasks with this command (I can see traffic on our monitoring of the upstream ports) … However, this task doesn’t show on the /toolkit/ page for the node or in a maddash throughput mesh.

 

I’ll have to check to be sure, but I think the only tasks that show up in the toolkit GUI are those that were set up there.  Ad hoc tests from the CLI don’t.

 

There is a handy “pscheduler monitor” command that will give you a real-time, top(1)-like display of what pScheduler is doing.

 

I feel that I’ve got something wrong in my pscheduler commands and that’s why the data isn’t being archived.  Does anyone have an example of a pscheduler CLI command that archives locally that they could share, please?

 

A better thing would be to run the task giving you trouble again, grab the run URL the CLI shows and hand it to “pscheduler result --archivings RUN-URL”.  That will dump out diagnostics for where each archiving went that include success or failure and any diagnostic information the archiver provided.  Your archive spec might be valid and the problem is being caused by something else.

 

--Mark