---

Subject:	Please share your regex and
Author:	[Not Specified]
Posted:	2014-02-10 11:34

Please share your regex and does the process you are monitoring use a specific tcp/udp port It might be easier to just monitor that instead.

Hydruid

---

Subject:	Same problem here
Author:	[Not Specified]
Posted:	2014-03-07 01:55

Did you ever find a solution for this I've got a new zenoss 4.2.4 install with SP405 applied. I can add a new OS process and as Paradox says modeling will detect the process, however killing the process will not generate an event. I've toggled the monitoring flag for the component on/off with no change.

regex: agent_manager
example: /usr/local/rvm/gems/ruby-1.9.3-p484@cloud_manager/bin/agent_manager

Any ideas This is a pretty big roadblock. Oh, I should also mention that I'm using SSH, not SNMP.

Thanks in advance

---

Subject:	On the custom OS process,
Author:	[Not Specified]
Posted:	2014-03-07 12:22

On the custom OS process, what did you set for the following values:

Enable Monitoring (zMonitor)
Failure Event Severity (zFailSeverity)

I would set them as follows:
zMonitor = Local Value Yes
zFailSeverity = Local Value Error

Hydruid

---

Subject:	Are you use that your regex
Author:	[Not Specified]
Posted:	2014-03-10 09:15

Are you use that your regex matches

Try this regex tester out: http://www.myregextester.com/

Hydruid

---

Subject:	Yeah, the regexes match
Author:	[Not Specified]
Posted:	2014-03-10 15:44

That's how the modeler detected them in the first place. I don't think I have a regex matching issue as far as I can tell. They are pretty basic. Zenoss just isn't detecting when a process goes away. On that note, it does appear to be at least occasionally detecting when this process restarts (I have restart notifications enabled).

Not sure how to detect whether or not it's doing a periodic check with the correct regex though...that might help me troubleshoot the issue.

---

Subject:	I'm guessing that something
Author:	[Not Specified]
Posted:	2014-03-11 10:35

I'm guessing that something with the regex is incorrectly detecting that the process is still running. Let me do some testing since I'm not a regex master and see if i can list some examples. I know the simplest option is to just put the process's name (e.g. java).

Hydruid

---

Subject:	Well, one of them is simply...
Author:	[Not Specified]
Posted:	2014-03-11 13:20

agent_manager. The other one I had to distinguish because it was also catching mongod :) The other simpler one behaves exactly the same. Do you know how to set the logging such that it will show all detection attempts with matches it finds/doesn't find

---

Subject:	Update
Author:	[Not Specified]
Posted:	2014-03-11 14:37

I'm not the pro in this area in Zenoss so I setup a simple test. I created a dummy process that I could stop/start easily. Below are the details and my results.

1. Created a new process called pingtest in the Zenoss Web UI. Set the values as:
zMonitor=yes
zAlertOnRestart=Yes
zFailSeverity=Error
Pattern=ping
2. Logged into a linux server and issued "ping 4.2.2.2"
3. Remodeled the server and boom got a new OS Process
4. Stopped the ping
5. The OS process never errored out.......very interesting

It turns out that if you're using the SSH template, OS Processes are discovered but aren't actually monitored. When using the SNMP template, OS Processes are discovered and are properly monitored.

Which template are you using to monitor the servers

Hydruid

---

Subject:	Additional Notes
Author:	[Not Specified]
Posted:	2014-03-11 14:38

The SSH template in Core requires modification to monitor many things, including interface utilization. By default, it only gives you a few basics.

Hydruid

---

Subject:	You really can't go wrong
Author:	[Not Specified]
Posted:	2014-03-12 15:16

You really can't go wrong with SNMP ;)

Hydruid