Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

RAID 1 Question

Status
Not open for further replies.

ITGUY53

MIS
Feb 6, 2003
62
0
0
US
hello-

i was wondering if someone can answer this fairly simple question. i already have an answer, but i wanted to make sure and concurr with people who's had more experience in RAID 1 issues.

i have a w2k adv server with a physical disk 0 & 1 set to RAID 1 disk image. i believe disk 0 is my primary where it is set with 2 partitions; C: (OS) & D: (Apps). my D: crapped out where check disk keeps wanting to run but i can't get through a certain stage and it eventually stays at stage 1 80%. so here's my question.

since my disk 0 & 1 is set to RAID 1 (disk image), could the currupted file(s) carry over to disk 1 from disk 0? my assumption is yes it has copied over as that's the job of RAID 1. please correct me if i am wrong or any othe input would be greatly appreciated.
 
Raid 1 is Mirrored images. Both physical drives are supposed to contain exactly the same data. There is no 'primary' drive. Is this a software raid, set up through the Operating system? I'm not sure how checkdisk would deal with that. If it's simply bad data, I would expect the data to be the same (bad or good) on both drives. If there's a physical problem, it could be only on 1 drive, but without some diagnostics, I'm not sure how you'd know which one has the problem. Most add on controller cards would tell you which one was bad. If this is a OS software created raid container, are there any diagnostics or reporting features beyond checkdisk?
 
the system RAID is not set in the OS level, but in the BIOS DELL PERC. i have checked the status of both drives on my server and they are both at OPTIMAL state. nothing wrong with either disk, and so this is strictly a software issue. they way these server was setup as RAID 1 probably didn't expect something like this to happen, but maybe only a hardware failure.

in w2k there are diagnostics and even tools to break RAID, but this isn't the issue. so i stand to be correct where RAID 1 completely just copies over exact data over to the other disk for a hardware setup. so is there is no point of taking out pulling out one drive because they are an entity of one, and two when you break the RAID 1 in hardware you will automatically lose the data? it would of been easier if it was set as RAID 5.
 
OK, now we're getting somewhere. If you have a PERC controller, the OS shouldn't even know about the raid system. First, make sure the PERC says that the drives are OK. You can do this through OpenManage (if set up) or check the status in the PERC bios. If the PERC says they are OK, it's almost certainly a Windows problem.

As for mirrored raid 1, if you remove 1 drive, the array will be degraded, but the data will not be lost. There will no longer be any redundancy (the R of raid), so if something else goes wrong at that point you'll be in trouble. When a drive is removed or (taken off line) and replaced, the array will need to be rebuilt - this might happen automatically or it might need to be done manually, depending on which PERC card & how it's configured.
 
so if something else goes wrong at that point you'll be in trouble."
Not quite, the disk you pulled is still an exact copy of the other disk, and can be used to create a new array, should the other disk die an untimely death (not considering the problem with the D partition)

"i can't get through a certain stage and it eventually stays at stage 1 80%."
If you have corrupt data on a raid 1, both disks have the same corrupt data. If a physical error appears on one of the disks, and the raid adapter is aware of it, the data involved in the damaged area is not copied over to the other disk in the set, nor is the physical error on one disk copied to the other disk. Not sure of check disk issue, but you may have to many physical disk errors for it to complete or you may have physical errors on both disks. Backup first, if your raid adapter supports patrol reads, run it. If not, I would break the mirror, run scsi tests on the drives individually, chkdsk the drive individually.

........................................
Chernobyl disaster..a must see pictorial
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top