[Check_mk (english)] Clustered Monitoring with Check_MK and DR:BD

Dear list,

I've one question to all of you: Does anyone of you have experiance with redundant monitoring where an OMD site is being clustered using
a shared storage with DR:BD? The idea is that you create let the
monitoring run on two servers, manage them with heartbeat/pacemaker
and us DR:BD as a common storage. Has anyone of you setup something
like that?

Greetings from Munich,

Mathias

···

--

Mathias Kettner

---
Mathias Kettner GmbH
Kellerstra�e 29, 81667 M�nchen, Germany
Registergericht: Amtsgericht M�nchen, HRB 165902
Gesch�ftsf�hrer: Mathias Kettner
http://mathias-kettner.de
Tel. +49 89 1890 435-0
Fax. +49 89 1890 435-29

Hi Mathias,

We had a drbd-based two-server-setup in the past with nagios and pnp (but not OMD or check_mk) on CentOS 5.3

The disk-io was almost too slow when using two dedicated gigabit-links for the drbd-traffic.

In the end we had about 1000 services - most of them with pnp graphs.

From time to time we also had split-brain issues with luci and ricci so we decided to do a clean re-install.

Our new setup is a standalone virtual machine on a 4-server ESX5-cluster using VMware HA.

We are now monitoring 113 hosts and 2250 services without any problems.

I would not recommend drbd, especially if there are lots of .rrd files to process.

best regards

Florian

···

Danke
&
mit freundlichen Grüßen

Ing. Florian Stichlberger

seteq systems
serving expert’s technology

+43 (0)512 318000
+43 (0)512 318000 8 (Fax)
info@seteq.at

http://www.seteq.at/


2013/3/25 Mathias Kettner mk@mathias-kettner.de

Dear list,

I’ve one question to all of you: Does anyone of you have experiance with redundant monitoring where an OMD site is being clustered using

a shared storage with DR:BD? The idea is that you create let the

monitoring run on two servers, manage them with heartbeat/pacemaker

and us DR:BD as a common storage. Has anyone of you setup something

like that?

Greetings from Munich,

Mathias

Mathias Kettner


Mathias Kettner GmbH

Kellerstraße 29, 81667 München, Germany

Registergericht: Amtsgericht München, HRB 165902

Geschäftsführer: Mathias Kettner

http://mathias-kettner.de

Tel. +49 89 1890 435-0

Fax. +49 89 1890 435-29


checkmk-en mailing list

checkmk-en@lists.mathias-kettner.de

http://lists.mathias-kettner.de/mailman/listinfo/checkmk-en

Dear list,

I've one question to all of you: Does anyone of you have experiance with
redundant monitoring where an OMD site is being clustered using
a shared storage with DR:BD? The idea is that you create let the
monitoring run on two servers, manage them with heartbeat/pacemaker
and us DR:BD as a common storage. Has anyone of you setup something
like that?

I asked this a few weeks ago and got an excellent
response:http://lists.mathias-kettner.de/pipermail/checkmk-en/2012-November/007768.html

I used http://blog.simon-meggle.de/tutorials/nagiosomd-cluster-mit-pacemakerdrbd-teil1/
(it's even in German for you!!) He did a follow up in english as well
with some other hints.

The one tweak I did was up the timeout for the omd status failover
(too sensitive, commits were bouncing the cluster). I changed it from
(i think) 20 to 90.

Patrick