Profound Logic Forums

We're having an issue where on occasion (about once a week or so) the sessions become unresponsive and this lasts for upto an hour or so. We have lots of users logged in and using the profoundUI sessions via the HTTP server. At this same time of the unresponsiveness, the 5250 sessions FLY. no issues at all. SO, I assume it has something to do specifically with the use of the HTTP servers but need some guidance on how best to troubleshoot this type of issue. Since moving to using ProfoundUI our inhouse skillset on dealing with this type issue is lacking.
Any ideas on troubleshooting steps?

Thanks

Occasionally I receive complaints that the sessions are running slow.

When checking the system, everything looks fine.

Usually, having the user completely close out of their sessions, closing their web browser & relaunching resolves the issue.

On other occasions, there are background scans (for malware, virus, etc) running on the users' machines, which slows down their machine in general.

I can't recall an instance (on my machine) where the http server was the issue. Then again, your setup may be more complicated than ours.

Thanks DanD

In our case, ALL the users get "locked up" for the HTTP sessions(profound). 5250 still hums along quickly.

I've not heard of this particular symptom before.

My guess would be that something is happening on your IBM i that is causing the HTTP server to slow down. I can't imagine its on the PC, because that wouldn't affect all sessions. It probably isn't the network, either, as that would affect 5250 and other things. So it must be something on the IBM i.

I would suggest that you take a look at how the QHTTPSVR subsystem is configured. Does it have sufficient memory available? What else is sharing the same memory pools? Are they running out of storage at the time when this occurs?

What about things like run priority? Is there something else running on the system at a higher priority when this happens?

Just in general I would look at what is happening on the system when the slowdown occurs... figure out what's happening that way.

None of this really has to do with Profound UI. It has to do with managing performance of jobs on the IBM i. You've noted that 5250 jobs are still performing well, you might want to look at the difference in the memory pools available to those, and the difference in run priority and other similar things.

Thanks Scott,
I'll take a look in that direction. I realize it's not specific to Profound, but our knowledge is with interactive 5250 so we were unsure where to even start looking.
We'll start looking into the memory pools etc. as suggested.