SQL Server Performance Forum – Threads Archive

Downside of Clustering?

Howdy We are considering a SQL Server 2000 cluster, and wanted to know what the reality was of running one.
A few questions to see what people think: (1) Which bits break ( hardware / OS / SQL Server ) (2) Just how relaible are they in the real world? (3) What configuration would you suggest and why? We can tolerate slower operation as long as everything is kept running.
Cheers,

Failover clustering increases server availability by allowing a system to automatically switch the processing for an instance of SQL Server from a failed server to a working server. With multiple instance support, you can isolate work environments (for example, testing from production) or volatile application environments and provide different system administrators for each instance of SQL Server on the same computer. 1) For a primary high availability solution, consider failover clustering. However, if you cannot afford to purchase the required complete clustered solution from the Microsoft Hardware Compatibility List (HCL), consider another high availability options such as log shipping, which requires HCL-compatible equipment, but does not require you to buy a complete solution or specialized hardware for a Windows cluster. 2) SQL Server 2000 Failover Clustering is the leading option as a primary method to achieve high availability for your databases. It offers full transactional consistency, and a failover to other another node that is automatic. By eliminating single points of failure both at the software and at the hardware level and with the proper processes and disaster recovery plans in place, you may be able to achieve five nines of availability.
3) For more informaition review [email protected]
http://www.sql-server-performance.com/clustering_intro1.asp
http://www.sql-server-performance.com/clustering_resources.asp
http://www.sql-server-performance.com/clustering_2000.asp Satya SKJ
Moderator
http://www.SQL-Server-Performance.Com/forum
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

Hi Satya, Thanks for your response. What I’m interested in is : * In real life support of clusters what fails on clusters and why -where are the weak spots – is it hardware/OS/SQL?? * What have you ( or others like Brad ) seen in real world support.
That sort of information is important to someone who is considering running a cluster. Cheers,

There are disadvantages with the Clustering :
– You incur a greater expense. The maintenance of two servers is two times the cost of maintaining a single server. Because you have to maintain two servers at the same time, it is more expensive to install and maintain clustered nodes.
– Servers should be in the same location. If the branches of the organization are across the globe and the Active/Active clusters must be implemented in the branches, the networking and the storage infrastructure that you have to use is very different from a standard quorum device server cluster.
– You have no protection against a disk array failure.
– Failover clustering does not allow you to create failover clusters at the database level or at the database object level, such as the table level. For more information about failover clustering, visit the following link:http://msdn.microsoft.com/library/en-us/architec/8_ar_cs_7ab7.asp"
Satya SKJ
Moderator
http://www.SQL-Server-Performance.Com/forum
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

Hi Sayta Thats great general info & I understand clustering basics as i have built them before…. What I’m really after is what you yourself have seen / experienced with clustering – you know…the 3am call outs for a broken quorum or an OS fault…that sort of thing. You can email me if you prefer… Cheers,

To be precise yes, I was called out for a failover issue during early hours which was happended due to Hardware fault and never had issues with the SQL application as such. Hope you get more information from Argyle in this regard. Satya SKJ
Moderator
http://www.SQL-Server-Performance.Com/forum
This posting is provided â€œAS ISâ€ with no rights for the sake of knowledge sharing.

Main issues I’ve had are: Shared disk array hardware or driver issues Bad Active Directory and/or DNS implementation making failover non successful. Comparing to a single server a cluster is great protection against "single" server issues. If one server crash you still have it running on the other one. But you still have a single point of failure which is the shared storage. Most issues I’ve encountered have been related to this. Issues with the shared storage can take a long time to solve and can require a total rebuild of the cluster. In that case it can be good to have a single server as stand by where you can restore your data until you rebuilt the cluster. On the other hand most issues with share storage will be encountered during setup. Once they are solved they usually tend to run without problems (until you apply a new service pack or patch <img src=’/community/emoticons/emotion-5.gif’ alt=’

‘ /> ). You just replace disks now and then. I’ve never had any issues with SQL Server itself that have been cluster related.

]]>