January 18, 2011 at 9:37 pm
Comments posted to this topic are about the item Creating A SQL Server 2008\2008 R2 Failover Cluster
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
January 19, 2011 at 3:16 am
If you're smart you'll make use of the new slipstream features and integrate the Service Pack and\or Cumulative Update to decrease the time taken to deploy your new SQL Server instance(s).
This is a must if you want to alter the collation from the US default. Otherwise you may need to rebuild the master database to alter the collation, not sure if this is still true in R2, but it certainly was before R2.
January 19, 2011 at 3:19 am
We're undergoing an enterprise consolidation and virtualisation project at the moment, and I currently have an external consultant proposing to replace our physical, clustered SQL Servers, with non-clustered VM's where the high availability is provided by VMWare HA.
Does anyone have a view on this? Any serious pitfalls / constraints to take into consideration?
Thanks in advance.
Reuben
January 19, 2011 at 5:38 am
Hello,
Can I receive a copy of the VMWare Server\SQL Server 2008 install.
Currently our site has a clustered 2008 R2 sql server instal built by a third party vendor and we are experencing huge network problems. The cluster has two nodes (SQL1,SQL2) with SQL2 being the active node and SQL1 being the failover node. We are recieving the following erros and user are unable to access the database at certain times of the day. We are at the mercy of the vendor o help us resolve the issue however they are not providing the proper resources to help us resolve the problem.
I would like to evaluate the Windows Server 2008 R2 64 bit and see if the installation has been performed correctly. Any help or guidance on this situation would be great.
Cluster network name resource 'SQL Network Name (XXXXXXSQL)' failed registration of one or more associated DNS name(s) for the following reason:
DNS signature failed to verify..
January 19, 2011 at 5:51 am
Nice article, however, I would include some reference to the MSDTC and SPN records, for instance:
"If you are installing only the Database Engine, the MSDTC cluster resource is not required. If you are installing the Database Engine and SSIS, Workstation Components, or if you will use distributed transactions, you must install MSDTC. Note that MSDTC is not required for Analysis Services-only instances."
Best regards ...
_Ak!
January 19, 2011 at 6:44 am
What edition of Windows & SQL does this article relate to? I would assume Windows Enterprise due to limitations on memory with Standard edition. Will this 3 node solution work with Standard SQL or does it require Enterprise?
January 19, 2011 at 7:30 am
BigSam (1/19/2011)
What edition of Windows & SQL does this article relate to? I would assume Windows Enterprise due to limitations on memory with Standard edition. Will this 3 node solution work with Standard SQL or does it require Enterprise?
Hi
as clustering requires a minimum of Enterprise for the operating system, I am using either Windows 2003 R2 Enterprise or Windows 2008 R2 Enterprise. SQL Server Standard edition supports 2 node clustering so this will be suitable for your test environment as you are only deploying each instance to 2 nodes.
INST1 nodes A and C
INST2 nodes B and C
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
January 19, 2011 at 7:33 am
Can I get the tutorial from you, please.
Thanks!
January 19, 2011 at 7:43 am
reuben.anderson (1/19/2011)
We're undergoing an enterprise consolidation and virtualisation project at the moment, and I currently have an external consultant proposing to replace our physical, clustered SQL Servers, with non-clustered VM's where the high availability is provided by VMWare HA.Does anyone have a view on this? Any serious pitfalls / constraints to take into consideration?
Thanks in advance.
Reuben
I believe that would mean you only have failover at the host level, not individual virtual servers.
---------------------------------------------------------------------
January 19, 2011 at 7:44 am
Perry,
maybe it is taken as read but it should be mentioned another drive would be needed to install SQL application files to (i.e. the instance root directory) and this would not be a cluster resource.
C drive could be used for this but a drive for the purpose would be usual.
thanks for putting the article together. Any plans for an article on clustering other SQL components (mainly SSIS and SSRS)
---------------------------------------------------------------------
January 19, 2011 at 7:58 am
how do we request a complete tutorial for the VMWare Server\SQL Server 2008 install ?
January 19, 2011 at 8:16 am
Just did one, don't make any mistakes, if you have any errors in initial install a pain to fix. Also understand the coponents, RS is not clusterable, you can put the back end db's there not the service.
January 19, 2011 at 8:22 am
Thanks for the great article....Would it be possible to get the complete tutorial for the VMWare Server\SQL Server 2008 install?
January 19, 2011 at 8:32 am
Hi, may I get the tutorial for the VMWare Server\SQL Server 2008 install please?
Thanks
January 19, 2011 at 8:32 am
reuben.anderson (1/19/2011)
We're undergoing an enterprise consolidation and virtualisation project at the moment, and I currently have an external consultant proposing to replace our physical, clustered SQL Servers, with non-clustered VM's where the high availability is provided by VMWare HA.Does anyone have a view on this? Any serious pitfalls / constraints to take into consideration?
Thanks in advance.
Reuben
plusses and minuses
the plus is that the failover is on vmware so if it does failover none of the SQL services stop. apps just won't be able to connect until it fails over to another node, but all the spid's will still be there. you can also ship the complete VM with all databases to a DR site and just mount it on another VMware node
another plus is that windows clusters are a pain in the *** and changing minor things like IP's can break them. no need for clusters on vmware
upgrading to new hardware is easier. you buy the new hardware and move the VM. no need to reinstall windows
the minus is that the I/O is still limited even if you dedicate drives to each SQL instance. if you have large databases with heavy I/O then you will have problems running them on one vmware node. all the I/O will go through the hypervisor
the db backups might also be a bit slow and you may have to switch to backing up the VM with a vmware aware backup solution like netbackup
the hardware issue can also be a minus. if you have a few SQL instances on a vmware box and you need resources that you don't have you will have to justify a purchase of an expensive server. with non-vmware sql you can buy a cheapo server and move a few databases from various servers to free up resources
Viewing 15 posts - 1 through 15 (of 57 total)
You must be logged in to reply to this topic. Login to reply