How to understand what loads VPS?

Question

linux

Hello. There are VPS. A couple of days ago this started:





Hosting started writing angry letters:

We notify you that on your VPS the Load Average value has reached a critical level, in connection with which the following processes have been stopped: sbin/licctl

Googled about the sbin/licctl process, it kind of refers to a request to use a license or something, and for many it also loads the CPU . I bought a license from a hoster, the situation immediately got better (this can be seen on the graphs).

Question number 1 - control of this process - whose responsibility is it, mine or hosting? Roughly speaking, there was no such jump before, can it be considered that the hosting has changed the configuration of VPS servers?

Question number 2. How to deal with the load further? Now the memory is full:



Here is the command

ps axo rss,comm,pid | awk '{ proc_list[$2] += $1; } END { for (proc in proc_list) { printf("%d\t%s\n", proc_list[proc],proc); }}' | sort -n | tail -n 10 | sort -rn | awk '{$1/=1024;printf "%.0fMB\t",$1}{print $2}'

produces the following:

746MB   metric
520MB   core
431MB   mysqld
180MB   php
109MB   apache2
93MB    cron
35MB    cron-ispmgr
34MB    fail2ban-server
33MB    sh
16MB    named

what is the metric process? Tried googling, couldn't find it.

Here is the situation for free -m:

total        used        free      shared  buff/cache   available
Mem:           2048        1637           0          86         410         138
Swap:             0           0           0

if anything, sorry for stupid questions, I don’t cook in administration))

In short, what should I do and who is to blame?

Answer 1

UPD:

Unfortunately, the panel found a problem that without a license, processes can accumulate. To avoid repeating the CPU problem, you can comment out the metrics rule in the cron job scheduler as follows:
###*/5 * * * * /usr/local/mgr5/sbin/cron-ispmgr sbin/metric >/dev/null 2>&1
To free the disk, delete the files in the /usr/local/mgr5/var/ispmgr_metric/ directory
An error in the panel operation has been registered. At the moment, we cannot specify how long it will take to resolve it. All errors are resolved in order of internal priority.

Answer 2

Googled about the sbin/licctl process, it kind of refers to a request to use a license or something, and for many it also loads the CPU. I bought a license from a hoster, the situation immediately got better (this can be seen on the graphs).

Oh, they are cunning...
According to the link, they issued a decision to stupidly remove an unnecessary service, if you don’t pay for it and don’t use it, why would it eat resources. ps if you have vps or dedicated, then the responsibility for everything that happens there hangs only with the owner, except when you use the system maintenance service (then you will not be given access to the
vps, which is correct) - this will be the software as service option
for example, he put you some kind of buggy software in his mind does not oblige him to anything. The car is yours, if you don't like it, delete it.

Answer 3

htop - similar to top, shows the current activity of processes
atop - a powerful process activity analyzer with file logging. - https ://firstvds.ru/technology/statistika-nagruzki...
you can record, for example, 1 minute / 10 minutes / 1 hour of activity time to a file and then watch every second or how you recorded all the activity of processes, who ate how much memory and processor, disk, etc.

Answer 4

ps aux|grep metric
It will show the path to the executable file, if anything, the output of the command to the studio :)

Answer 5

it's some sort of rip-off.
what kind of license?
just use the top command to see the load.
if st is high, hosting, or rather neighboring VPS steal your CPU - the node is overloaded. It's not your fault, but the host's.
If your software slows down - demand a transfer to another node or dump it to another host.