Issue High CPU psaadm in

[daanse]

Hi,
i often have High CPU in the Night.
How can i eliminate that Problem?

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
29891 psaadm    20   0  275980  48776  36228 R 100,0  0,2   0:00.16 sw-engine
29892 psaadm    20   0  279024  51636  35588 R 100,0  0,2   0:00.16 sw-engine
28986 gl_user   20   0  502928  98552  64672 R  97,3  0,4   2:48.29 php-fpm
29889 psaadm    20   0  279224  52036  35776 R  97,3  0,2   0:00.15 sw-engine
29890 psaadm    20   0  279024  51892  35836 R  97,3  0,2   0:00.15 sw-engine
29893 psaadm    20   0  279224  52540  36280 R  97,3  0,2   0:00.15 sw-engine

Plesk Onyx
Debian 8.x

 

[Bitpalast]

During the past months there have been several threads on a high cpu usage by sw-engine. It seems that Plesk engineers need to look into that issue.

For your do-it-yourself approach try to run
# ps aux | grep sw-engine
while the cpu load is high. It will display more details on what scripts sw-engine is running. From these scripts it should be able to derive the cause. Could you please post that process list output when you got it? So far no other use has posted it, so when new threads alike this one are opened, we're all stuck with a solution, because noone has went further then to show the "top" list. We'll need more details to find the cause of the high load.

 

[daanse]

Uff, okay, if i had more details, i would provided it. ;-)
Problem is, it was 2 o' Clock in the Night
And it seems only to be for a short time.
I will keep you updated.

@Peter Debik i think i got notified through Watchdog.
The top output is from Email.
Is there a way to tune Watchdog, to send me also other outputs like ps aux?

In this Case my Server is doing this several Times a Week.

 

[Bitpalast]

If this happens only once at night and at approximately the same time it is caused by either a backup or other daily maintenance tasks. This can normally be ignored.

 

[daanse]

  PID USER      PR  NI    VIRT    RES    SHR S  %CPU %MEM     TIME+ COMMAND
10705 psaadm    20   0  279240  52172  35736 R 100,0  0,2   0:00.16 sw-engine
10702 psaadm    20   0  279240  51944  35684 R  95,7  0,2   0:00.15 sw-engine
10704 psaadm    20   0  282312  55780  36372 R  95,7  0,2   0:00.15 sw-engine
10706 psaadm    20   0  277688  50552  35776 R  95,7  0,2   0:00.15 sw-engine

i had this Problem again.
And i ran the command.
Its not clear to me, can you see something conspicuous?

 

[Bitpalast]

Please refer to https://support.plesk.com/hc/en-us/articles/213919445-Process-sw-engine-produce-high-CPU-usage

 

[daanse]

Hi,
thankyou, that's awesome.
But i think i am missing something. Actually i have Plesk Mobile Center 1.7-4, is this not the latest Version?

 

[Bitpalast]

I have posted a question to that FAQ entry, because the same problem and same question came up recently for another Plesk user. It is obvious that the push-worker script is causing the issue, but I do not know why it is and what the solution can be. Maybe it is a bug, but unclear at this point. If you have the opportunity to open a support ticket with Plesk, that would be the way to go from here. Else we'll both need to wait if someone reviews the FAQ article and comments on it.

 

[WebHostingAce]

I had a similar issue. It was caused by Plesk failing to restart Apache gracefully.

Issue with Apache Adaptor.

Have a look in Apache Error Log and see if you can notice Apache failing a gracefully restart and going for a second full restart.

Support from @Peter Debik and Plesk Team, i was able to solve that issue.