Issue sw-engine status red

[Gernot Schabkar]

hello.
i got 2-3 times a day the cpu alert red for about 20 minutes. this is a fresh server with about 12 abonnements , 15 Domains, 8 cpus and 16gb ram. that shouldnt be?

30023 psaadm 20 0 273232 45664 35708 R 100.0 0.3 0:00.21 sw-engine
30022 psaadm 20 0 273540 46228 35828 R 98.4 0.3 0:00.20 sw-engine
30024 psaadm 20 0 273532 45932 35632 R 98.4 0.3 0:00.20 sw-engine
30025 psaadm 20 0 273248 45504 35524 R 98.4 0.3 0:00.20 sw-engine
30026 psaadm 20 0 272908 45124 35340 R 98.4 0.3 0:00.20 sw-engine

what is going wrong, where to search for solutions?

 

[Gernot Schabkar]

I counted, i got 8 mails a day concerning CPU load with all colors a day.
all with the psaadm user.
the server is doing close to nothing every day , only 5 webhosts. i planned putting 500 on this server, thats not looking good...

Message:
Der Status des Serversystem-Parameters "Services > CPU-Auslastung durch Apache" wurde von "yellow" in "red" geändert.

top - 09:37:23 up 48 days, 18:43, 0 users, load average: 0.15, 0.55, 0.51
Tasks: 222 total, 6 running, 216 sleeping, 0 stopped, 0 zombie
%Cpu(s): 4.1 us, 0.1 sy, 0.0 ni, 95.7 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st
KiB Mem: 14402584 total, 13929152 used, 473432 free, 432620 buffers
KiB Swap: 62499836 total, 74116 used, 62425720 free. 11940844 cached Mem

PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
30023 psaadm 20 0 273232 45664 35708 R 100.0 0.3 0:00.21 sw-engine
30022 psaadm 20 0 273540 46228 35828 R 98.4 0.3 0:00.20 sw-engine
30024 psaadm 20 0 273532 45932 35632 R 98.4 0.3 0:00.20 sw-engine
30025 psaadm 20 0 273248 45504 35524 R 98.4 0.3 0:00.20 sw-engine
30026 psaadm 20 0 272908 45124 35340 R 98.4 0.3 0:00.20 sw-engine

 

[Bitpalast]

Are you using the Plesk Mobile Extension?
https://support.plesk.com/hc/en-us/articles/213919445-Process-sw-engine-produce-high-CPU-usage

 

[Gernot Schabkar]

in the link, http://autoinstall.plesk.com/extensions/packages/ , the latest seems to be Version 1.4 , i´ve got version 1.7 installed. so u suggest to downgrade?

 

[Bitpalast]

No, not suggesting a downgrade. It was just one possibility for what is causing the high sw-engine load.

 

[Gernot Schabkar]

not getting better with mobile disabled.

 

[Bitpalast]

Do you or customers have any crontab jobs active ("geplante Aufgaben")? Which ones?

I would like to annotate that 16 GB RAM is far too small for the planned 500 domains. I suggest to upgrade to at least 64 MB unless you want to constantly use a swap file. However, this only as a note, it is not causing the sw-engine load.

 

[Gernot Schabkar]

agreed, i can scale it up to 128 gb.
no, nothing in the crontabs yet.
maybe i should mention, that i had tomcat causing some huge cpu loads in the past, but i didnt need it, so i stopped the service.

 

[Bitpalast]

Your best chance to find out what is causing this is to run
# ps -aux | grep sw-engine
while the problem exists. As the issue occurs for a longer period of time, it should be possible to catch the window and look at the process list. It will display more details what commands sw-engine is running at that time. These will probably allow a better insight into what is happening.

 

[Gernot Schabkar]

maybe another hint, or another problem..
im getting every some days the error in my backoffice :
Fehler: Der ModSecurity-Regelsatz konnte nicht aktualisiert werden: DB query failed: SQLSTATE[42S02]: Base table or view not found: 1146 Table 'psa.Configurations' doesn't exist, query was: DESCRIBE `Configurations`

 

[Bitpalast]

That's no good. It means that there is a potentially severe problem in your Plesk database setup. Very hard to remotely guess what it is and why it happend. According to this article it happened during an upgrade:
https://support.plesk.com/hc/en-us/...Column-not-found-Base-table-or-view-not-found
I am unable to give correct advice on this, because in my opinion it involves too many variables.

 

[Gernot Schabkar]

is that procedure messing my existing customers on this server?
i installed it on a fresh maschine ...

 

[Bitpalast]

I am sorry that I cannot assess the future behavior of the system in this case. There is an obvious defect in your psa database. It can have a severe negative impact on the installation, but it does not necessarily have to.

You can try to
# plesk repair db -y
to fix it. That is pretty safe, because psa database is auto-backed-up during the repair, so it could be re-imported in case things go totally wrong.

Else if you feel unsafe with instructions given in FAQ article opening a support ticket is a good option, too.

 

[Gernot Schabkar]

plesk repair db -y

Checking the Plesk database using the native database server tools .. [OK]

Checking the structure of the Plesk database ........................ [OK]

Checking the consistency of the Plesk database ...................... [OK]

Error messages: 0; Warnings: 0; Errors resolved: 0

didnt do anything..
btw. have many thanks for your help.