Question about running a second instance | SQL Server Performance Forums

SQL Server Performance Forum – Threads Archive

Question about running a second instance

We have two Windows 2003 servers running one instance of SQL Server Standard Edition on each Windows server. Both servers have two 2.79GHz Xeon processors with 2.5 GB of ram. What we are thinking about doing is putting a second instance on each Windows server to contain “standby/backup” databases of the other server. In other words we have server A and server B, with two databases on one instance on each server and a backup/standby of server A#%92s databases on the second instance of server B and a backup/standby of server B#%92s databases on the second instance of server A. If either server A or B goes down we would have minimal downtime and the developer would only have to change the connection string. To me this seems like a no brainer, since this is sort of a disaster plan and the case where either server would have both instances (four databases) actively being used would be a short amount of time, hopefully days at most. But he wants my recommendation on weather this is a good idea before he starts coding for this scenario. I have searched the web for information about second instances of SQL server but most of the information I#%92ve found on second instances is when people have two versions of SQL server installed. Please post any experiences, bad or good, about second instances of SQL Server Standard Edition. Thanks

Are you going to replicate these servers? CanadaDBA
Both servers are production server?
Luis Martin
Moderator All postings are provided “AS IS” with no warranties for accuracy.
If the hardware is good enough to resource these 2 instances and no issues for the application performance then I don’t see any issue in installing second instance on both the boxes. If they want it for warm standby then use Log shipping in order to keep the databases in sync. HTH Satya SKJ
This posting is provided “AS IS” with no rights for the sake of knowledge sharing.
As long as you are not overly concerned with performance than this would be an acceptable DR strategy. If you could invest in a 3rd stand alone DR box, this would provide the same functionality in a disaster scenario but not affect the performance on the main production database. I would use log shipping to synchronize the two databases but that is a personal preference. There are several ways to do this. -Jason
Both servers would be production.I do a restore of the DB nightly. I don’t do log shipping because I’m concerned with constantly monitoring the restore of the trans logs and the possibility of the trans restores getting out of sync. It takes about 10 minutes to restore all the trans logs, so that is not a real concern. I am concerned with performance. The users/developer will not accept more that a second or two response time. Thanks to everyone for your input.