I am creating custom distributed application in SCOM 2012 R2. We want to monitor the whole Lync 2013 infrastructure. We have 2x front end server role + 2x mediation server role + lync edge server + reverse proxy server + SQL archiving database + office web apps server

I am wondering what is the best approach to build such distributed application, I mean what objects\classes should be used within it? E.g. when I define Lync servers component like below using LS server role class it means that when e.g. the disk space will be almost filled out it will not influence the state of such distributed application. Am I right?

So what is the best practice in such situation? What objects should I use to monitor e.g. Office web apps within such distributed application?

Some blogs may be useful for you, please refer to the below link

Configuring Lync Server 2013 to work with System Center Operations Manager

http://technet.microsoft.com/en-us/library/jj205188.aspx

one Lync and SCOM thread:

https://social.technet.microsoft.com/Forums/lync/en-US/1ff30e7d-a2b6-4828-b42c-e534fb587c41/monitoring-lync-with-scom?forum=ocsmanagement

Hi bbelko,

the short answer to "what to add to the DA" is.. "it depends".

rather than add "too much" I would take the approach of "what is critical to me?" or "what do I really need to know?".

I would meet the Lync administrator, ask him/her what does he/she wants to know or is critical in the infrastructure and then implement that in the DA. However, an agreement must be done in terms of "how much is enough". This can be really granular, and the application can easily grow complex. That's why I would start with something very basic, then start to increase more objects, etc.

hope this helps!

Jose

The problem is that actually I am one of Lync administrators and I need to decide what is most important for Lync infrastructure. What I found in a meantime is "Health Rolllup Diagram" module in Lync server 2013 Management Pack Guide, part of it below:

A Lync Server pool can contain multiple individual Lync Server systems (with more than one Lync Server server role, Lync Server service, and Lync Server component). Therefore, the failure of an individual server or component is less critical to the overall health of the Lync Server pool, because other servers in the same pool can provide the application service to the client. The health will roll up on a percentage level to the Lync Server pool.

The Lync Server Pool Watcher performs synthetic transactions against a Lync Server pool. Consecutive failure of one or more synthetic transactions (a process known as the consecutive polling interval) will roll up the critical health state to the pool level (worst of any synthetic transaction), as shown in the following diagram.

Note: The Lync Server Port Watcher and global Simple URL Watcher do not roll up the health state.

Using this information I think that I should start with using LS site class for monitoring Lync servers (Mediation and FrontEnd), am I right? But not sure what I should do with the rest of infrastructure, SQL server, Office web apps and the rest

Hi bbelko

Im struggelig with the exact same challenge, to decide what objects/classes to add to the DA.

Have you been able to design a DA for your lync deployment yet? If so could you share your design in this thread?

Regards
Jonatan

Not really, I did not have enough time to do it so it is not done yet