12-03-2006, 10:55 PM
"Yes I do. I have had them crash, updates go bad... running them on a virtual server makes them much easier to restore. "
Okay...I'll buy that at the slightest bit on both server types and the updates. Depends on your backup solution(s) though. I'll stick with what I do. Guess I am just luckiest of the lucky and none of the (not) many updates have 'gone bad'.
"You're right, your method would use the hardware more efficiently. However, the overall efficiency goes way down. When the OS becomes corrupted or fails after an update, it takes MUCH more work. The cost of added hardware to run a VS at the same speed, is easily surpassed by the cost of brining a machine back up to speed."
I disagree...not much work at all; should a complete failure happen - utilizing another reason below (redundant systems).
"Let's say there's 100 users on a machine. Each have their files and dev work saved on that server. If it's using your method and a user uses his admin access to erase all files on the machine, you're SOL. If that user was on a Virtual Server, the other 99 users wouldn't be affected."
Why would standard, authenticated users have admin rights on these machines? Properly set up (even applications that ahere such), they wouldn't.
Let's be realistic. Yes, you would have a few/several of these servers setup to properly balance the load...even an initial switching server re-directs them to ServerX or ServerB...but even in a 30,000+ campus...very, very doubtful that ~100 users...let alone ~30 are going to be hitting one of these servers for application usage at any one time - personally speaking.
"The largest Windows solution provider in the west uses Virtual Servers for just about everything."
And that is? Thankfully...never had to utilize a Windows solution provider.
Okay...I'll buy that at the slightest bit on both server types and the updates. Depends on your backup solution(s) though. I'll stick with what I do. Guess I am just luckiest of the lucky and none of the (not) many updates have 'gone bad'.
"You're right, your method would use the hardware more efficiently. However, the overall efficiency goes way down. When the OS becomes corrupted or fails after an update, it takes MUCH more work. The cost of added hardware to run a VS at the same speed, is easily surpassed by the cost of brining a machine back up to speed."
I disagree...not much work at all; should a complete failure happen - utilizing another reason below (redundant systems).
"Let's say there's 100 users on a machine. Each have their files and dev work saved on that server. If it's using your method and a user uses his admin access to erase all files on the machine, you're SOL. If that user was on a Virtual Server, the other 99 users wouldn't be affected."
Why would standard, authenticated users have admin rights on these machines? Properly set up (even applications that ahere such), they wouldn't.
Let's be realistic. Yes, you would have a few/several of these servers setup to properly balance the load...even an initial switching server re-directs them to ServerX or ServerB...but even in a 30,000+ campus...very, very doubtful that ~100 users...let alone ~30 are going to be hitting one of these servers for application usage at any one time - personally speaking.
"The largest Windows solution provider in the west uses Virtual Servers for just about everything."
And that is? Thankfully...never had to utilize a Windows solution provider.