HX Node CPU Usage After Upgrade

[bhackbarth]

I just changed an HX controller to 6.0 PR5 last night (because of DTMF dialing issues).  It's working well, but I noticed the CPU usage is pegged constant at 35% (user) with occasional spikes to 50%. Normally, when I see this on a 5000, I backup the database, test and repair it, then upload the tested/repaired database even if it said nothing needed to be repaired.  Just wondering if this is the next step or if there is another recommendation?  The controller itself is working.  I did do a Reset System from DB Programming this morning before everyone arrived but that didn't clear it.

[acejavelin]

I see this kind of CPU usage all the time, even in smaller systems... Why do you think this is an issue if they are not having problems?

[bhackbarth]

I thought it's an issue because I don't normally see it; this behaviour stood out. And, it never did that on PR2.   After hours, I did do the "backup, test/repair, restore" trick and now the cpu idles along at 2% like before and just like all my other 6.0 sites with similar numbers of users.

When it was pegged yesterday it took forever (or at least longer than normal) to access DB Programming, get voicemail email attachments, etc.

[Tech Electronics]

bhackbarth,

Some upgrades do a restructuring of the files and the CPU will be under load while it is fixing the actual data that is on the Flash Drive to match the new structure. You doing a backup/restore will do that as well, but it shouldn't be necessary. Give it about 30 minutes and see if it resolves itself. If it doesn't then do your backup/restore to speed things along if you think it is necessary; besides it never hurts anything to do that anyway.

Thanks,

Steven