CPU usage configuration report in cervlet

[Martin Pala]

Hi,

I have three little issues:

1.) if monit is configured to report CPU usage, it shows in cervlet 
process configuration listing "actual_cycles/max_cycles" count, as follows:

Max. CPU usage     If greater than 2.0% for 0/5 cycle(s) then alert

You can see above, that there is "0/5". To the console mode (monit -v) 
monit outputs:

CPU usage limit      = if greater than 2.0% for 5 cycle(s) then alert


What tabout to synchronize it? I think it will be better to not report 
actual cycles count, so cervlet will give:

CPU usage limit    If greater than 2.0% for 5 cycle(s) then alert

(which is exactly what is defined in monitrc)


2.) what about making start+stop programs in passive mode optional? 
Currently if they aren't defined, monit won't monitor the service. I 
think it will be usefull to allow passive mode monitoring live without 
start+stop definition. It will be needed also to print information in 
the case that star/stop program wasn't defined  in cervlet and terminal 
mode if user will try to start/stop process via monit's httpd or by 
"monit start".


3.) Last question is regarded to cevlet and terminal process information 
order - what about to move mailinglist information to the end of process 
info? If you few contacts, important informations such as protocol test 
are at the end of the list, which as maybe not good. I think that we 
should move all these informations above process mailing list, so we can 
have for example:


Process Name          = calendar-dwp
Monitoring mode      = passive
Group                = (not defined)
Pid file             = 
/global/ldapims/d1/ics/SUNWics5/cal/bin/config/pidfile.dwp
Start program        = /usr/bin/true
Stop program         = /usr/bin/true
Host:Port            = 192.168.100.105:9779 [protocol DEFAULT]
CPU usage limit      = if greater than 2.0% for 5 cycle(s) then alert
Every                = (not defined)
Timeout              = Do timeout if 5 restart within 5 cycles
Alert mail to        = address@hidden
  alert from         = (default)
  alert subject      = (default)
  alert message      = (default)
  alert on timeout   = yes
  alert on restart   = no
  alert on checksum  = no
  alert on resource  = no
  alert on stop      = no
  alert on timestamp = no
Alert mail to        = address@hidden
  alert from         = (default)
  alert subject      = (default)
  alert message      = (default)
  alert on timeout   = yes
  alert on restart   = yes
  alert on checksum  = yes
  alert on resource  = yes
  alert on stop      = yes
  alert on timestamp = yes

instead of:

Process Name          = calendar-dwp
Group                = (not defined)
Pid file             = 
/global/ldapims/d1/ics/SUNWics5/cal/bin/config/pidfile.dwp
Start program        = /usr/bin/true
Stop program         = /usr/bin/true
Host:Port            = 192.168.100.105:9779 [protocol DEFAULT]
Timeout              = Do timeout if 5 restart within 5 cycles
Alert mail to        = address@hidden
  alert from         = (default)
  alert subject      = (default)
  alert message      = (default)
  alert on timeout   = yes
  alert on restart   = no
  alert on checksum  = no
  alert on resource  = no
  alert on stop      = no
  alert on timestamp = no
Alert mail to        = address@hidden
  alert from         = (default)
  alert subject      = (default)
  alert message      = (default)
  alert on timeout   = yes
  alert on restart   = yes
  alert on checksum  = yes
  alert on resource  = yes
  alert on stop      = yes
  alert on timestamp = yes
CPU usage limit      = if greater than 2.0% for 5 cycle(s) then alert
Every                = (not defined)
Monitoring mode      = passive