This ends up reading a lot like story problem you'd see in school, but I am at the end of even beginning to guess what the issue could be here. ANY new suggestions would be highly appreciated!
We have a Dell 2900 that has 8 HD caddies on the front. (Controller 0 on the top and Controller 1 on the bottom on a Perc 5/i card) We bought it with controller 0 full of Seagate 73gb 15k drives in a RAID 5 setup. The primary use of this machine was going to be for some light (4ish) Virtual machine use. The machine itself is a 2 quad core Xeon 3.0ghz with 16gb of RAM.
One of the Virtual machine is going to be using a lot more HD space than we originally thought so we purchased 5 refurbished Fujitsu 143gb 15k drives to put in the Controller 1 spot and use that specifically with the virtual machine that required the space.
Aaand here's where it gets crazy:
When we plugged in the drives (called HD1-5 here) into the proper slots (named 1:4-1:7) we got a problem where the drives would all start up and the two lights on the caddies would be green, and then after a few minutes two of the drives (most often slot 1:6 and 1:7) would go into a "degraded state". I don't even remember what the possessed us to try it, but we pulled the HDs out of slot 1:4 and 1:5 and rebooted it. Slot 1:6 and 1:7 light up green and staid green. The drives were no longer marked as degraded.
So we thought it might be the HDs so we swapped them around and no. As long as there are 4 or 3 drives in Controller 1 only 2 of them will work/ be recognized by the system.
At that point we went on an update binge and updated the firmware on EVERYTHING. Controller card, backplane, HDs, everything we could get our hands on. No change. It still would only recognize two drives plugged in there are once. (When there are only 2 drives in (it doesn't matter which HDs they are, any of the 5 act the same way) Dell open manage says everything is fine. Put 3 or 4 drives in and suddenly the backplane is listed as "degraded"... for BOTH controllers. However that status doesn't effect the performance of Controller 0 at all.)
At this point we called the company we bought the server from (not Dell) and they said, "Oh, that has GOT to be the controller card. We'll send you out a new Perc 5/i card to replace it." Great. We get the card and install it.
We boot up the machine and every drive errors. We repair the "foreign" RAID setup that is on Controller 0 and then check the drives in Controller 1. It still only sees two of them. We check firmware again, we rearrange the drives. Same problem. Ok, so it wasn't the controller card.
We call up the company again and we're told then it's probably the backplane. They send out a backplane (and we requested all new cables to make sure that wasn't an issue as well) and replaced all that when it came in. Fire up the machine and we have the SAME issue. 3 or 4 drives and the backplane gets listed as "Degraded". 1 or 2 drives (it doesn't matter which slot it's in) and the backplane is no longer "degraded" and everything is fine.
Some other things I have tried: RAIDing two of the new drives in a non "degraded" state and then rebooting to see if the 3rd and 4th drive would then show up. No luck. It just degrades the Virtual Drive so it's unusable.
We also checked the power supply and did some math on it and the (two) power supplies are each 950W. Even with some really extreme numbers I can't imagine that power would be an issue.
That is just about it I think. Any new suggestions and or ideas for things to try would be appreciated.
Thanks...
We have a Dell 2900 that has 8 HD caddies on the front. (Controller 0 on the top and Controller 1 on the bottom on a Perc 5/i card) We bought it with controller 0 full of Seagate 73gb 15k drives in a RAID 5 setup. The primary use of this machine was going to be for some light (4ish) Virtual machine use. The machine itself is a 2 quad core Xeon 3.0ghz with 16gb of RAM.
One of the Virtual machine is going to be using a lot more HD space than we originally thought so we purchased 5 refurbished Fujitsu 143gb 15k drives to put in the Controller 1 spot and use that specifically with the virtual machine that required the space.
Aaand here's where it gets crazy:
When we plugged in the drives (called HD1-5 here) into the proper slots (named 1:4-1:7) we got a problem where the drives would all start up and the two lights on the caddies would be green, and then after a few minutes two of the drives (most often slot 1:6 and 1:7) would go into a "degraded state". I don't even remember what the possessed us to try it, but we pulled the HDs out of slot 1:4 and 1:5 and rebooted it. Slot 1:6 and 1:7 light up green and staid green. The drives were no longer marked as degraded.
So we thought it might be the HDs so we swapped them around and no. As long as there are 4 or 3 drives in Controller 1 only 2 of them will work/ be recognized by the system.
At that point we went on an update binge and updated the firmware on EVERYTHING. Controller card, backplane, HDs, everything we could get our hands on. No change. It still would only recognize two drives plugged in there are once. (When there are only 2 drives in (it doesn't matter which HDs they are, any of the 5 act the same way) Dell open manage says everything is fine. Put 3 or 4 drives in and suddenly the backplane is listed as "degraded"... for BOTH controllers. However that status doesn't effect the performance of Controller 0 at all.)
At this point we called the company we bought the server from (not Dell) and they said, "Oh, that has GOT to be the controller card. We'll send you out a new Perc 5/i card to replace it." Great. We get the card and install it.
We boot up the machine and every drive errors. We repair the "foreign" RAID setup that is on Controller 0 and then check the drives in Controller 1. It still only sees two of them. We check firmware again, we rearrange the drives. Same problem. Ok, so it wasn't the controller card.
We call up the company again and we're told then it's probably the backplane. They send out a backplane (and we requested all new cables to make sure that wasn't an issue as well) and replaced all that when it came in. Fire up the machine and we have the SAME issue. 3 or 4 drives and the backplane gets listed as "Degraded". 1 or 2 drives (it doesn't matter which slot it's in) and the backplane is no longer "degraded" and everything is fine.
Some other things I have tried: RAIDing two of the new drives in a non "degraded" state and then rebooting to see if the 3rd and 4th drive would then show up. No luck. It just degrades the Virtual Drive so it's unusable.
We also checked the power supply and did some math on it and the (two) power supplies are each 950W. Even with some really extreme numbers I can't imagine that power would be an issue.
That is just about it I think. Any new suggestions and or ideas for things to try would be appreciated.
Thanks...