Hi,Actually, I have an all-in-one SCOM R2 server (SCOM, SQL, Reporting) with 8Gb ram. I don't have many Views or Dashboard for now. Some operators will be using the console for monitoring. I am wondering if it is better to deploy the Web Console on this server or having a console deployed on each user's desktop ? How much will be the load (memory / cpu) on the server for five to ten operators ? Is there any metrics to calculate that ?Thanks,Alain

Hi.SCOM R2 has a view which shows how many connections to the SDK service are made. Every SCOM Console connection (whether webbased or not) needs a SDK connection in order to work. So the ammount of SDK connections reflects in a kind of way how many Console connections are running. Having said that, putting it all on one box is only advised for testenvironments. Not for production since you will end up with a slugish SCOM environment or one that totally freezes. So my advice, outside your question is to move the SCOM DBs over to a dedicated server. Makes a huge difference.But back to your question now :)Whatever Console is used, it creates a load on the RMS. A web based Console even more since that component is mostly installed on the RMS as well in order to skip the Kerberos Double Hop issue. In that case IIS puts another load on the RMS as well. Otherwise the Web Console needs to be installed on a non RMS with the usage of formbased authentication OR the server has to be configured for delegation.Woa! That's a lot of information. Whenever you need more explanation please let me know.So basically when you choose between the Web Console or the full blown version is more a choice of WHAT functionality is needed. For an Operator the Web Console is good enough (mostly) but for a SCOM Admin the full blown version is needed. Best regards, Marnix Wolf (Thoughts on OpsMgr)

Alain,I think Marnix is exactly right here. The load should be pretty comparable, so focus on features / usability differences between the two. I'd opt for the OpsConsole myself.RE: Hardware / load etcBe sure to take a look at our Sizing Helper to more guidance on the hardware reccomendations based on the Number of Agents monitored etc. Its always goot to have more disk / memory resources, but without enough performance will suffer.http://www.microsoft.com/downloads/details.aspx?displaylang=en&FamilyID=b0e059e9-9f19-47b9-8b01-e864aebf210cMichael Pearson OpsMgr Performance Test Team http://blogs.technet.com/michaelpearson/ This posting is provided "AS IS" with no warranties, and confers no rights. Use of attachments are subject to the terms specified at http://www.microsoft.com/info/cpyright.htm

Hi, Just to say Thank you for your appreciated feedback.Alain

Marnix, How are you making this determination? I see the exact opposite in my testing of the # of SDK connections. I have two web console servers and I see the # of connections increase for each of them, but I have 3 Web Consoles open to each server the the SDK connections count didn't increase for the additional consoles that were opened. Also from my desktop when I open the web console I don't have an SDK connection to the RMS only the Web Console server does. Can you recheck this and tell me if you get the same or let me know what I'm missing? This begs the question does the Web Console really have the same impact as the Full Console on the SCOM mgmt group? I can't see how it does if we are strictly going by SDK connections. Thank you Marc Schmieder

If you have Citrix, use that for running the "fat" console. This will centralize management (ie: only one place to put CUs) as well as allow remote users to use the console with no lag. We're using it here and it works great."Fear disturbs your concentration"

We don't have Citrix in place and that is a good solution for using the console in remote sites, but it doesn't answer the question around SDK connections or if each Web Console actually uses an SDK connection as each fat client does. I'm looking for someone to disprove or reaffirm what I said previously. thank youMarc Schmieder