Re: Question about check priority

Hello Guillaume,

Yes I can see this is one right now.

Process 'unicorn_master_myapp'

status OK

monitoring status Monitored

monitoring mode active

on reboot start

pid 21645

parent pid 1

uid 1000

effective uid 1000

gid 1000

uptime 1d 2h 37m

threads 2

children 3

cpu 0.0%

cpu total 19.1%

memory 1.1% [175.5 MB]

memory total 52.5% [8.0 GB]

security attribute unconfined

disk read 0 B/s [1.4 MB total]

disk write 0 B/s [1.1 GB total]

data collected Thu, 26 Aug 2021 16:49:13

Process 'unicorn_myapp_worker0'

status OK

monitoring status Monitored

monitoring mode active

on reboot start

pid 7017

parent pid 21645

uid 1000

effective uid 1000

gid 1000

uptime 17h 11m

threads 3

children 0

cpu 0.6%

cpu total 0.6%

memory 2.2% [343.0 MB]

memory total 2.2% [343.0 MB]

security attribute unconfined

disk write 0 B/s [15.8 MB total]

data collected Thu, 26 Aug 2021 16:49:13

Process 'unicorn_myapp_worker1'

status OK

monitoring status Monitored

monitoring mode active

on reboot start

pid 5580

parent pid 21645

uid 1000

effective uid 1000

gid 1000

uptime 1d 0h 26m

threads 4

children 0

cpu 0.6%

cpu total 0.6%

memory 2.1% [331.7 MB]

memory total 2.1% [331.7 MB]

security attribute unconfined

disk write 0 B/s [45.3 MB total]

data collected Thu, 26 Aug 2021 16:49:13

Looking at logs now :

Aug 27 15:27:02 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' mem amount check succeeded [current mem amount = 190.7 MB]

Aug 27 15:36:32 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' process is not running

Aug 27 15:36:32 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' trying to restart

Aug 27 15:36:32 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' process is not running

Aug 27 15:36:32 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' trying to restart

Aug 27 15:36:54 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' process is running with pid 32304

Aug 27 17:10:21 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' mem amount of 358.9 MB matches resource limit [mem amount > 350 MB]

Aug 27 17:10:21 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' exec: '/bin/bash -c /bin/kill -TERM `cat /var/run/service/unicorn_myapp_worker1.pid` && sleep 1'

Aug 27 17:10:21 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' mem amount of 358.9 MB matches resource limit [mem amount > 350 MB]

Aug 27 17:10:21 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' exec: '/bin/bash -c /bin/kill -TERM `cat /var/run/service/unicorn_myapp_worker1.pid` && sleep 1'

Aug 27 17:10:51 ip-172-31-32-81 monit[10922]: 'unicorn_myapp_worker1' mem amount check succeeded [current mem amount = 184.5 MB]

It looks it is being killed right after matching the limit not respecting the 8 cycles. defined as individually below

```

check process unicorn_master_teachworks

with pidfile /var/run/engineyard/unicorn_myapp.pid

start program = "/bin/bash -c ‘/service/bin/app_myapp start'"

as uid deploy and gid deploy

stop program = "/bin/bash -c ‘/service/bin/app_myapp stop'"

as uid deploy and gid deploy

if mem > 350 MB for 2 cycles then restart —> Is this group directive overwriting the individual one below?

if cpu > 100% for 2 cycles then restart

group unicorn_myapp

check process unicorn_myapp_worker1

with pidfile /var/run/service/unicorn_myapp_worker1.pid

if mem > 350.0 MB for 8 cycles then exec "/bin/bash -c '/bin/kill -TERM `cat /var/run/service/unicorn_myapp_worker1.pid` && sleep 1'"

group unicorn_myapp

```

Thank You,

Carlos Cano

On 26 Aug 2021, at 07:27, Guillaume François <guillaume.francois55@gmail.com> wrote:

Hello,

According to your configuration, it should be 3 different Monit processes.

Do you see 3 elements while issuing command "monit status" ?

Also the log file should contain more details about what is detected and which actions are performed from monit.

Regards.

Le jeu. 26 août 2021 à 02:48, <hcano@solucoes.net> a écrit :
Hello,

I am using Monit version 5.25.1

My config is:

```
check process unicorn_master_myapp
with pidfile /var/run/services/unicorn_myapp.pid
start program = "/bin/bash -c '/services/bin/app_myapp start'"
as uid deploy and gid deploy
stop program = "/bin/bash -c '/services/bin/app_myapp stop'"
as uid deploy and gid deploy
if mem > 350 MB for 2 cycles then restart
if cpu > 100% for 2 cycles then restart
group unicorn_myapp

check process unicorn_myapp_worker0
with pidfile /var/run/services/unicorn_worker_myapp_0.pid
if mem > 350.0 MB for 8 cycles then exec "/bin/bash -c '/bin/kill -TERM `cat /var/run/services/unicorn_worker_myapp_0.pid` && sleep 1'"
group unicorn_myapp

check process unicorn_myapp_worker1
with pidfile /var/run/services/unicorn_worker_myapp_1.pid
if mem > 350.0 MB for 8 cycles then exec "/bin/bash -c '/bin/kill -TERM `cat /var/run/services/unicorn_worker_myapp_1.pid` && sleep 1'"
group unicorn_myapp
```

Unicorn has a master process then each worker has another.

In this case if the master process is above 350MB for 2 cycles it has to be restarted.

Now the workers has a limit of 8 cycles.

My question is:

I am seeing workers being killed after 2 cycles.
Is it intentional as the check on group unicorn_myapp is in some kind overwriting the later ones?

What I may be missing?

Thank You,
Carlos Cano

From:	hcano
Subject:	Re: Question about check priority
Date:	Fri, 27 Aug 2021 15:53:46 -0300