Site resilience is a requirement of a lot of companies but on a lot of occasions it is not a requirement to achieve full local resilience immediately following a site failure. In other words a lot of companies are deciding that in the event of the loss of a single data centre site the service would resume in a second data centre but using only a fraction of the kit and might run the core services only. After the service has resumed a decision would need to be made as to whether to establish full service and resilience in the 2nd data centre or choose to fail back...
So the easiest way to achieve this with Exchange 2007 and the mailbox role is to have your SCR target as a single server that could be built out as a cluster in the event of site failure. So is this technically possible and supported? The answer to both of these questions is yes ...but with some caveats.
There is now a lot of great information on high availability in general here: High Availability
More specifically this section (Standby Continuous Replication: Site Resilience with Standby Clustering) provides a run through of the processes involved in activating an SCR target. The steps are based on Exchange 2007 running on Windows 2008 but there are references to Windows 2003 where relevant.
Of course, as this section of Technet does touch on, 'High Availability' should apply to more than just the mailbox role and in my experience designing to reach your required levels of high availability for client accessibility, message routing etc etc. can be just as challenging...
PingBack from http://www.biosensorab.org/2008/02/11/single-node-cluster-as-scr-target/
Site resilience is a requirement of a lot of companies but on a lot of occasions it is not a requirement
I read that the SCR target does not have to be a single node cluster, it can also be standlone server ?
What are the advantages to each ?
I am planning to deploy SCC with a SCR.
I think the answer to this question depends on what your second data centre is for. Is your second data centre ever going to be used permanently or is it only ever going to be used temporarily whilst the issue with the main data centre is rebuilt. For some companies the secondary data centre is a temporary thing where only a subset of services are run from in the event of the loss of the primary site. In this case it might make sense not to provide the same levels of resilience in the second data centre ..therefore use a standalone server as the SCR target. If the second data centre is designed only to be used permanently in the event of a catastrophic event taking the primary data centre out of action then it makes sense to deploy an equivalent level of resilience as the primary ..therefore use SCC as the target.
Whichever you decide it's worth walking through the steps for activating the SCR target with your chosen solution because this might sway your decision...
>Whichever you decide it's worth walking through the >steps for activating the SCR target with your chosen >solution because this might sway your decision...
Well said. Test it ! ;-)
i need to know the validity of using the target SCR as a single passive cluster node , and the Quorum should be Local Quorum ? , then run the "setup.com /RecoverCMS /CMSName:CCREX01 /CMSIPAddress:172.16.6.153" cmdlet
so need to know the CMSIP should be in the DR site or in the same subnet of the CCR site .
if valid the whole way to recover the CMS using one cluster passive node
..the IP address can be on a different subnet and the target can be in a different AD site. The node will have a local quorum until you invoke DR. At this point you use /recovercms as you say and you recover the service on one node. Your users can log in an this is supported. At some point in the future you need to add the second node to restore resilience.
Hope that helps.
thx for helping ..
when going to run the setup.com /RecoverCMS command , it was showed that the cluster should have two NIC card on the server as the CCR server , so one will assigned in the public network and the other will be private untill the second node ?!!!!
The best practise is recommended to separate out cluster communication obviously. Since you only have one node this isn't an issue anyway but during cluster setup you should see warnings. Are you seeing warnings when you run /recovercms?
This might help? http://technet.microsoft.com/en-us/library/aa997910.aspx
i'm about using the feature of manual database seeding instead of WAN seeding, i've dismount the database and copy the .edb as a flat file and paste it in its target ( SCR ) machine, then i had run resume-storagegroupcop cmdlet, but i found that there were many logs in the copy queue lenth and did't apply on the database...
I have an SCR target with a single NIC and im getting the warning that the node has fewer than 2 NICs configured. If I install a second NIC, what network should it be connected to??