We have a completely virtual Windows infrastructure running on ESX and use
SCOM 2007 for all of our monitoring purposes. We don't do any monitoring of
the VMWare kit using SCOM, but there is a management pack available (from
Nworks)

SCOM does an excellent job of letting us drop the number of consoles we were
monitoring. Previously we had Dell IT Assistant, Citrix console, Mutiny
network monitor and a whole host of other consoles to look at. With SCOM and
all the relevant management packs all of the alerts are fired to a single
console, so all we need to monitor is the SCOM console and the VI client. As
I am just one guy supporting and monitoring sixty servers as well as
designing new infrastructure, I find this invaluable - I just look up and
see if a server is low on disk space, reporting predictive hardware failure,
or simply exceeding a performance threshhold.

I find that the default settings for many of the MS management packs give
you utter overkill on your problems, but I find that I'd rather know
everything that's going on than nothing. I have ironed out lots of
superficial problems that otherwise I would never have known about from
these messages. For instance, the OCS management pack told me that some of
my telephone numbers in AD were in the incorrect format. You can easily
override or disable monitors if the information is extraneous. One thing
that does annoy me is that you can't seem to stop monitoring for a scheduled
time period, but that's a minor concern.

The worst part I find about SCOM is tuning it to provide a good quality of
alerts to other operators. Give them too much, and they start to ignore
them. Give them too little, and problems slip by. I am quite happy to sift
through heaps of alerts and pick out the vital ones, but others don't have
the patience.

However, it is great for a single console for management of your windows
infrastructure. I broke the database on it recently and was lost without the
visibility it gives me. If a service stops, you get an alert telling you it
has failed and a link to restart it. The days of writing scripts to monitor
all sorts of server health events are long gone, and you certainly don't
have to jump from system to system trawling event logs. It is also quite
straightforward to set up your own management packs. I quickly and easily
designed a little bit of monitoring to tell me when my WebSense licenses are
going to run out, and it auto-starts a script to reset the WebSense services
which saves me from having to take any action when it occurs.

The bigger the infrastructure and the bigger your support staff, however,
means more of a headache with setting it up and updating it. I find it an
absolutely invaluable tool, but YMMV

I am currently trying to get SCCM working as well, but the arrival of VDI
means that this has been put on the backburner for a while.

Cheers,


2009/3/6 Ziots, Edward <[email protected]>

>  We are having Microsoft coming in to talk with us about Systems Center
> for Management of our ever-growing server farm, for those using it, please
> feel free to give me your pro’s con’s and tales from the trenches on what
> this platform said it will do, and what it really does when the rubber hits
> the road.
>
>
>
> My Mix of servers are ½ physical ½ virtual, SQL (2000/2005), IIS (5&6),
> File and Print (2000-2008), DC’s (2003), Application servers ( 3rd party)
> (2000/2003)
>
>
>
> TVK, I especially want to hear your thoughts on this subject, since this is
> your MVP realm and you tend to know the most about the features and
> functionality.
>
>
>
> TIA in advance,
>
> Z
>
>
>
> Edward Ziots
>
> Network Engineer
>
> Lifespan Organization
>
> MCSE,MCSA,MCP+I, ME, CCA, Security +, Network +
>
> [email protected]
>
> Phone:401-639-3505
>   ------------------------------
>
>
>
>
>
>

~ Finally, powerful endpoint security that ISN'T a resource hog! ~
~ <http://www.sunbeltsoftware.com/Business/VIPRE-Enterprise/>  ~

Reply via email to