• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Project: Rackmount Overkill

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Whats the amount per VM?
That answer depends on what I'm running in it. At this point, I could give each VM 24gb each and still have 24 left over for the host OS.

Most of my CentOS clients are 512mb to 1g. Windows "desktop" versions are 1g-2g depending on what they are doing. Windows server versions are 4g+.
 
IMO id say until the avg per VM falls under 1-2GB its overkill
Wouldn't the average VM be getting larger?

Also, if it wasn't clear to anyone reading the thread, I ordered 192gb of RAM for my server. Twelve of these, to be specific: Samsung 16GB PC3-12800R (DDR3-1600) ECC registered, Model M39B2G70BH0-CK0
 
You said you can do 24GB to each VM and host now, so until that shrinks to below whatever is acceptable to you, then Im sure its fine.


How much does that motherboard support?
 
Darn, you saw the other thread. Here is 196gb of ECC Registered server RAM.

196gb.JPG

The server is prepping for stability testing as I type this. I'll let it run for a bit to make sure everything is set and ready to go. I did have one stick that the board refuses to work with, so I'm going to RMA it. Speaking of which, this board has an interesting feature. It tests all the RAM modules individually at startup and isolates any broken/unstable ones. When it finishes the POST, it tells you exactly which stick is having problems. Never had a board do this before and it saved me a ton of troubleshooting time.
 
Darn, you saw the other thread. Here is 196gb of ECC Registered server RAM.

The server is prepping for stability testing as I type this. I'll let it run for a bit to make sure everything is set and ready to go. I did have one stick that the board refuses to work with, so I'm going to RMA it. Speaking of which, this board has an interesting feature. It tests all the RAM modules individually at startup and isolates any broken/unstable ones. When it finishes the POST, it tells you exactly which stick is having problems. Never had a board do this before and it saved me a ton of troubleshooting time.

The engineer who designed that is going straight to heaven. :shock:

Is this going to your Opty server? What are you going to do with the sixty four gigs of extra RAM? Or is this for your PEs?
 
Yes, this is going into the G34 server. I'll probably sell off the old RAM. I was going to use it in a system, but it is ECC memory, which wouldn't work very well.
 
Started at 1.45pm and it is currently 8.01pm. It has not completed a single iteration of the test.
 
Ok, so you know wanted to know how long it takes with that big of a matrix? The answer is:

Screenshot from 2012-10-25 23:22:53.png


29204 seconds
486 minutes
8.1 hours

"a lot"

Just started Memtest to run overnight.
 
Server is updating now before everything goes live again.

In the 17 hours it was running memtest, it did 1.75 passes.
 
Server is back up, fully functional. VirtualBox was throwing fits since it updated.

Screenshot from 2012-10-26 19:18:49.png

Time to fill that RAM bar up.
 
Back