Comments on: How do you do Disaster Recovery http://www.allenkinsel.com/archive/2010/01/how-do-you-do-disaster-recovery/ SQL Server, PASS, and other data mishaps Thu, 22 Jul 2010 13:53:20 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: Allen Kinsel http://www.allenkinsel.com/archive/2010/01/how-do-you-do-disaster-recovery/comment-page-1/#comment-65 Allen Kinsel Mon, 01 Feb 2010 18:37:53 +0000 http://www.allenkinsel.com/?p=134#comment-65 Funny you should mention a virtual DR Farm, I believe that we're doing something similar, but the problem as I see it is how do you physically get the servers recovered? regardless of the hardware, there has to be a better method of getting the software up. Since most DR sites arent permanent having everything running fulltime offsite isnt an option Funny you should mention a virtual DR Farm, I believe that we're doing something similar, but the problem as I see it is how do you physically get the servers recovered? regardless of the hardware, there has to be a better method of getting the software up. Since most DR sites arent permanent having everything running fulltime offsite isnt an option

]]> By: Brent Ozar http://www.allenkinsel.com/archive/2010/01/how-do-you-do-disaster-recovery/comment-page-1/#comment-60 Brent Ozar Fri, 29 Jan 2010 21:56:34 +0000 http://www.allenkinsel.com/?p=134#comment-60 One of the slickest ways I've seen recently is to have a completely virtual DR farm. Run just a few VMware (or Hyper-V) hosts, and let each team build up their own virtual servers in DR. You have limited hardware horsepower in DR, but you only do testing of apps there, not full blown production. You might not even have enough horsepower to turn on all of your DR VMs at the same time. Then when disaster strikes, you can easily grab more VMware host hardware, spread the load across more servers, and add capacity easily. You don't have to buy hardware before you need it, and you can usually beef up the hardware fairly quickly even in disaster situations. It's much easier to do this than to build the servers from scratch, and users are tolerant of decreased performance rather than having no servers at all. One of the slickest ways I've seen recently is to have a completely virtual DR farm. Run just a few VMware (or Hyper-V) hosts, and let each team build up their own virtual servers in DR. You have limited hardware horsepower in DR, but you only do testing of apps there, not full blown production. You might not even have enough horsepower to turn on all of your DR VMs at the same time.

Then when disaster strikes, you can easily grab more VMware host hardware, spread the load across more servers, and add capacity easily. You don't have to buy hardware before you need it, and you can usually beef up the hardware fairly quickly even in disaster situations. It's much easier to do this than to build the servers from scratch, and users are tolerant of decreased performance rather than having no servers at all.

]]> By: Pat Wright http://www.allenkinsel.com/archive/2010/01/how-do-you-do-disaster-recovery/comment-page-1/#comment-59 Pat Wright Fri, 29 Jan 2010 04:05:10 +0000 http://www.allenkinsel.com/?p=134#comment-59 What about mirroring? to another SQL Server somewhere else? Is that an option? Also what kind of time do you need to be back up in? That usually helps to determine some of your options as well. What about mirroring? to another SQL Server somewhere else? Is that an option? Also what kind of time do you need to be back up in? That usually helps to determine some of your options as well.

]]>