Hi People
There is one Stretched Farm setup I’m gonna involve in. Yes, I did good enough research on it. I know <= 1ms && >= 1Gbps is kind of difficult/impossible in WAN scenarios. However, I have a few queries regarding the implementation of it.
1. Is there any end-to-end implementation guide for setting up Stretched Farm? the only post that mentioned some steps is HERE
2. In general we execute Configuration wizard to add a server to existing farm
How it’s going to be in Stretched Farm scenario?
3. I have another thought that it may not be about adding each individual servers in the Secondary datacenter to the existing farm – is it true? if so, how is it different from DR farm?
4. Some of the servers may fail at times (or) down. How the fail over scenarios are handled? – I read redundancy is answer for this, however I wanted to know how the request is routed to the redundant resource(server).
5. Coming to Database layer, in most of the posts it was a database mirroring between Primary & Secondary datacenters. However, with new versions of SQL Server mirroring may not be supported. any other approach for this?
Thanks
Sarat
Thanks @Aniket. Now it’s my turn to implement it. 🙂
I had done this recently for our client. I have merged 2 SharePoint 2010 farm into one and upgrade it to 2013.
we have 2 Datacentre with dark pipe between them and we have this <= 1ms && >= 1Gbps available over WAN.
The Environment:
 Data centre A : 1WFE,1 APP server, 1 APP server Bound to search app only, 1 OWA server, 1 DC Server, 1 SQL  Server 2014 standard version.
UPS is running on this Data centre only.
Data centre B: All of above in Datacentre A  in addition to File server which is also going to be acted as Witness server for mirroring and SQL DB Backups.
2. In general we execute Configuration wizard to add a server to existing farm. How it’s going to be in Stretched Farm scenario?
– You need to specify all details when you run the config wizard and it will identify the config DB and add the servers to the farm
3. I have another thought that it may not be about adding each individual servers in the Secondary datacenter to the existing farm – is it true? if so, how is it different from DR farm?
– It will be only one farm but the servers are spread across 2 data centres.
– DR farm is the another farm than the existing farm. So, in general you have 2 farms if you have DR Farm.
4. Some of the servers may fail at times (or) down. How the fail over scenarios are handled? – I read redundancy is answer for this, however I wanted to know how the request is routed to the redundant resource(server).
In our Farm The WFE servers are load balanced using third-party load balancer. So if the Datacentre A goes down users are able to access the sites as all the load goes to the Datacentre B.  We need to switch over some Services like UPS to Datacentre but that is not critical.
5. Coming to Database layer, in most of the posts it was a database mirroring between Primary & Secondary datacenters. However, with new versions of SQL Server mirroring may not be supported. any other approach for this?
We have SQL server 2014 standard version and we are currently using mirroring on our farm and it is working like a charm. It has been failed over to the mirror during the working hours and users even didn’t noticed any changes from their end. One of my SQL expert friend said it is there in SQL 2016 but for that you need to test out as i have not done anything in SQL 2016
Hope this will help you.
