Re: monit 4.6 problem

[Pavel Urban]

Pavel Urban wrote:
> Hello,
>
> monit is acting strangely on our Solaris 8 server.
>
> At 1:51, test of LDAP server failed. Probably log rotation, database 
> maintenance or something like that, it is quite normal. Two cycles later 
> it was ok, BUT - monit keeps sending out alerts and finaly unmonitors, 
> even when in logs it says everything is ok. Well, it's not exactly true 
> - it says that is OK over and over again, which is not normal, either.
>
> Is it known problem?

Further exploration revealed that it is probably different problem.

I wanted this: when some important service fails, send me an email 
alert. Repeat sending for some time, then send me a sms and stop 
monitoring. Don't restart or anything.

I've tried to do it this way:

set alert address@hidden with reminder on 10 cycles
set alert address@hidden on { timeout }

check process ldap-master with pidfile 
global/ldapims/d1/ldapmaster/slapd-master-1/logs/pid
  if failed host 192.168.100.107 port 389 protocol ldap3 then alert
  if failed host 192.168.100.107 port 389 protocol ldap3 for 5 cycles 
then unmonitor
  mode passive

This configuration behaves as following: when it detects problem, it 
sends out warning.

Jan  4 01:51:19 ims1 monit[5358]: [ID 702911 user.error] LDAP: error 
receiving data -- Resource temporarily unavailable
Jan  4 01:51:19 ims1 monit[5358]: [ID 702911 user.error] 'ldap-master' 
failed protocol test [LDAP3] at INET[192.168.100.107:389].

After some (quite short) time, monit detects recover and send out 'pass' 
alert.

Jan  4 01:55:36 ims1 monit[5358]: [ID 702911 user.error] 'ldap-master' 
connection passed

After cca 18 minutes however, it sends out some kind of 'unmonitor' 
alert, but without actually unmonitoring or logging anything (note the 
last line)

*********************
Connection failed Service ldap-master

        Date:   Wed, 04 Jan 2006 02:09:39 +0100
        Action: unmonitor
        Host:   ims1

Your faithful employee,
monit

'ldap-master' connection passed
*********************

From this time, monit keeps sending this mail for every 20 minutes.

???

--
***********************************************************************
Pavel Urban (address@hidden)
IOL system disaster
Internet OnLine, www.iol.cz (owned by Czech Telecom, www.ct.cz)
***********************************************************************
   Vegetables should not operate electronic equipment.
          Computer Stupidities, http://rinkworks.com/stupid/
***********************************************************************