9.1 upgrade - HIGH cpu usage from java

[jggbbg]

We're having the same trouble here as well. x64 version 9.1 on a quad Xeon. CPU usage was low before upgrade. Now it lives on medium to high and bombs out from time to time. We can't have down time as online testing begins at our school system this week.

[mrunkel]

Please contact support. I think this has been said about a million times so far. But please, contact support. There is no "patch" for this issue as the causes have been very different.

I'll say it again. Please contact support.

http://www.untangle.com/live-support

We are extending support to all users, so please, take advantage of it.

[jggbbg]

I would have loved to contact support first thing this morning, but I'm on the East coast and have to wait until you guys open.

Once I was able to get support on the phone, the support tech ran a postgres patch on our server. CPU usage has dropped significantly; down to 2-4 as opposed to 15-27! I just hope the increase in CPU usage and server locking up were the same issue (or related).

If it goes down during testing this week I might be burned at the stake by our faculty and staff.

[JCVieira]

I'm having the same problem since the 9.1 upgrade. I'm using 9.2 now and the problem persists. Installed VM tools, did also allocated one more CPU, tried everything but I still cannot surpass the problem.

[wirepuller134]

Seeing the same issue here at random times, I thought it was a hardware issue until I saw this post. We installed external monitors for temperature on the chassis, processor and harddrive. All temperatures have been steady and in what I would call the normal range.
Just posting to let them know someone else is seeing it as well. We are seeing it at random times once or twice a week. Network load has been at zero and at times very heavy. And yeppers we will be contacting support and thank you for that!!

HP slimline PC (approximately 5 years old, it was retired from desktop use)
Sempron 3600+ processor
Nvidia 6100 graphics
3 gigs DDR 5300
Sata 250GB western digital drive
Secondary network card has an Intel chipset, don't remember the exact model.
Untangle version 9.1.1

[WebFooL]

JCVieira,

I would start a own thread there you post so much info as you can.
EX what VM Hardware version, Tool version hardware configuration both on the virtual VM and the ESX/ESXi box.

Then we will be able to help you better.

And welcome to the forum!

[Mathiau]

seems my box is at it again high spiking CPU usage from the GUI monitors apparently with 15min avg of 6-7 , dropping all traffic it seems except things i have in bypass rules, like internal company IM, i cant event RDC into my network or access the GUI from outside.

trying to see what it may be, this has been happening i am told every night around the same time, but seems past when reports run...

[Rhino Box]

Mathiau, what version and architecture of Untangle do you have? This has been fixed APPARENTLY on version 9.2, we have ALL of our boxes with this version, even 9.2.1, and we have not seen again the issue. Let me know.

Also the CPU usage for this issue, we saw that the HHD was the one that was really being the problem, since the reading and writing tasks were literally over the limit, so it may be something regarding the databases purge and HHD IOR, or something like that.

Even so, I would recommend if you are a paying customer, to write to support mail system to see what they can do.

Best regards and luck on this issue.

[Mathiau]

i am on 9.2 lenny, didnt do the 9.2.1 upgrade yet, i am out of the country at the moment but i am back today and plan to push out the 9.2.1 update tonight, was going to send in a ticket but of course when you need a hotels internet to work properly it doesnt..lol and figured if i tried to push the update out something would go wrong and then i would be really screwed!

it does have an older SATA 80G HD in it, my new UT box is sporting a Intel 320 SSD so hopefully that will help also for any issues as i do like to try and keep logs for more than 7 days as i do reviews every 2 weeks or more, so keeping data for only a week doesnt work well for me.