February 23, 2017 at 9:15 am
Hello, I have a client that wishes to transform our existing single node system into a high availability environment. I have never worked on a system with clustering and have not set one up before. We are implementing our system in the AWS GovCloud right now. I have 1 instance already set up and configured in single node mode. Do I have to reinstall the existing instance from scratch or can we just stand up another server, install it as a cluster and then connect the other one to the clustered server?
Also, I need to put together some pros and cons of doing this scenario. We also have the option of presenting a COLD failover scenario or even a MEDIUM level. I assume doing failover clustering adds a lot of administration overhead ... also how does storage work? Does each server have to have the same drive set up and would it take up the same amount of space on both instances?
Any help and guidance is greatly appreciated. 🙂
February 23, 2017 at 9:29 am
You will need to redeploy that Instance as a Clustered instance, you can't convert it.
Failover clustering has a big massive central point of failure: shared storage.
You could alternatively look into availability groups, using standalone instances, but requires dedicated storage, and a quorum of sorts (can be a file share witness on a central server) to each instance to host a copy of the databases in your AG.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
This thing is addressing problems that dont exist. Its solution-ism at its worst. We are dumbing down machines that are inherently superior. - Gilfoyle
February 23, 2017 at 9:38 am
Thank you,
So in failover clustering, both servers would "share" the same disk drives from one of the servers?
February 23, 2017 at 10:02 am
You'll need central storage, not dedicated to either.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
This thing is addressing problems that dont exist. Its solution-ism at its worst. We are dumbing down machines that are inherently superior. - Gilfoyle
February 23, 2017 at 10:03 am
amy26 - Thursday, February 23, 2017 9:38 AMSo in failover clustering, both servers would "share" the same disk drives from one of the servers?
The shared disk drives would be on a SAN or some sort of external shared storage, not in one of the servers that's a node in the cluster. You could use local disks in the servers to hold the TEMPDB (SQL Server 2012 and newer).
The administrative overhead of using clustering is mainly in getting the OS clustering setup, and getting all the resources and dependencies of the SQL Server cluster setup properly in Failover Cluster Manager tool. Once it's setup the first time, it's fairly easy to move the services to other nodes, specify a preferred node, etc. As Henrico mentioned, Availability groups has an advantage over clustering since it's not dependent on the shared storage. I haven't worked with AG so I can't say very much about it.
February 23, 2017 at 10:05 am
Availability Groups... is that the new Always On stuff?
February 23, 2017 at 10:14 am
Yes, it is a part of it. Here's some articles about the topic:
http://www.tech-coffee.net/alwayson-availability-group-introduction/
https://www.brentozar.com/sql/sql-server-alwayson-availability-groups/
https://msdn.microsoft.com/en-us/library/ff877884.aspx
https://technet.microsoft.com/en-us/library/hh510230(v=sql.110).aspx
I'm sure there's tons more.
February 24, 2017 at 5:21 am
amy26 - Thursday, February 23, 2017 9:15 AMHello, I have a client that wishes to transform our existing single node system into a high availability environment. I have never worked on a system with clustering and have not set one up before. We are implementing our system in the AWS GovCloud right now. I have 1 instance already set up and configured in single node mode. Do I have to reinstall the existing instance from scratch or can we just stand up another server, install it as a cluster and then connect the other one to the clustered server?Also, I need to put together some pros and cons of doing this scenario. We also have the option of presenting a COLD failover scenario or even a MEDIUM level. I assume doing failover clustering adds a lot of administration overhead ... also how does storage work? Does each server have to have the same drive set up and would it take up the same amount of space on both instances?
Any help and guidance is greatly appreciated. 🙂
see my stairway to AlwaysOn on this site at this link
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 8 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply