Unsolved
This post is more than 5 years old
2 Posts
0
74173
Replacing SATA disk in PERC S100 RAID 1 configuration
I have a Windows 2008 server (32-bit, T110) with two 500GB SATA disks in a PERC S100 RAID 1 config. One of the disks is showing signs of failure so I am going to replace it. OpenManage is installed if I need it.
What I would like to do is shut the server down, take a full disk image backup, then replace the disk.
- What tools can I boot from to image the disk (to a spare 1TB SATA drive)?
- I assume all I need to do to replace the drive is to take the old one out and put the new one in?
- Roughly how long will it take for the new disk to be synchronised with the remaining old one?
I'm looking to keep downtime to a minimum, and I need to travel to site to do this so want to make sure I have everything I need with me.
DELL-Chris H
Moderator
Moderator
•
8.4K Posts
0
November 5th, 2014 06:00
Hollymcr,
Just to clarify, when you said that it is showing signs of failure, are you getting a Predicted Failure showing on that drive? I ask as the process will be different if so. You have to account for the predicted failure before rebuilding. Now in regards to tools needed for booting outside of the OS, with the S100 you won't be able to do this. The S100 requires the driver from 2008 in order to run. So your best choice will be backing up the data from within the OS and then implement the rebuild. The tools we would offer (OMSA Live) is Linux based, which doesn't work with that controller.
In regards to the rebuild time, it will depend on a couple things. The Virtual Disk size, as well as current load on the server. I would think it will take about an hour or so.
You can find more information on the S100 here if needed- http://ftp.dell.com/Manuals/all-products/esuprt_ser_stor_net/esuprt_dell_adapters/poweredge-rc-s100_User%27s%20Guide_en-us.pdf
Let me know if this helps answers your questions.
hollymcr
2 Posts
0
November 5th, 2014 09:00
There is no predicted failure showing. However Windows event viewer shows several recurring faults on disk 0:0 which subsequently clear, but where this used to be a once-per-month thing it's now anything up to several per minute, hence the decision to replace it.
Re backups: Presumably, if I image the raw disk before I start (using a sector-by-sector copy), then although I may not be able to access the files from another machine I should still be able to restore the disk to the starting point if anything goes wrong?
Is there any reason why booting something like Acronis from a BartPE disk wouldn't be able to load the right drivers?
Otherwise, how would you suggest I carry out a full system backup directly from within the OS?
rawnet
1 Message
0
December 25th, 2014 21:00
my experience is that acronis does not support perc s100/300 controllers.