Three times more CPU load after upgrade

Yesterday I upgraded from 2.0.11 and asterisk 11 to FreePBX 13 and asterisk 13. After this the load consuming by the asterisk process seems 3x higher then before the update.

Is this normal? How can I check this?

from the root console, “top” and hit enter and post the top 10 lines or so - it will show what is eating the CPU.

Yeah, it’s the asterisk process, here 280%

top - 14:13:50 up 1:48, 4 users, load average: 0.04, 0.20, 0.30
Tasks: 212 total, 1 running, 210 sleeping, 0 stopped, 1 zombie
Cpu0 : 26.3%us, 35.2%sy, 0.0%ni, 38.1%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu1 : 31.4%us, 34.1%sy, 0.0%ni, 34.1%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu2 : 28.6%us, 32.8%sy, 0.0%ni, 38.2%id, 0.0%wa, 0.0%hi, 0.4%si, 0.0%st
Cpu3 : 30.4%us, 30.8%sy, 0.0%ni, 38.7%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 8051600k total, 734436k used, 7317164k free, 19928k buffers
Swap: 1738748k total, 0k used, 1738748k free, 124540k cached

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
6511 root 20 0 2793m 170m 13m S 280.4 2.2 19:44.20 asterisk

What does the console look like? On an idle system (may have to wait for Idle) there is almost nothing being sent to the console except for the housekeeping routines - I wonder too if you have a looped channel somewhere - again, when idle, do a core show channels - if it is truly idle, there should be no channels active.

When idle there is no problem.

It starts when we start making calls. On the old server (Freepbx 11 and Asterisk 11) there is no problem handling 20 or more calls, load rarely goes over 50%.
After the update we see CPU use hitting 400% with only a few calls active.

What Processor are you using - is this a Pi?

No, not a PI.

It’s a virtual server with 4 Intel E5 cores assigned to it.

Then something is way wrong with that VM - Here is a picture of one of my hosts:

All Hyper-V VM’s, all with 4 Processors and 4 Gig’s of RAM - I never see more than 2-3% on the CPU’s - even when I am doing maintenance.

What are you using for Virtualization?

I don’t think it’s VM related. I can simulate the problem on two different platforms. Also the problem occurs only with Asterisk 13. Asterisk 11 ran multiple years without any problems.

We are running about a 50-50 mix right now of 11 and 13 - A little over 100 machines - and we see no difference in CPU - And all the machines in the Hyper-V capture above are running 13. It’s got to be something specific with your setup.