HP DL 380 G4 server problems with SCSI hard disks

[DragosPD]

Hello, I have problems with my DL380 G4 server: The failure led's on hard drives goes amber and the drives are disabled.

I tried reseating the backplane, SCSI cable, PCI controller, everything. I used two controllers: the default one from the motherboard and the smart array p600 controller.

I have six 72GB hdd's at 15k rpm. I configured the array like this: two of them in RAID 1 with one spare the other two also in raid 1 with one spare.

I have updated the firmware to the latest release.

I mananged a few weeks ago after restarts/reseating internal parts/changing between available controllers to mount suscessfully all six drives. Yesterday when I copied a lot of files in the server (so I heavily used the hard drives/controller/file system) my linux server crashed cause five of the hard drives goes amber and disabled.

My questions are:
1. What part's did I need to change? The SCSI cable? The backplane? The controller could be the problem - despite that I have two?

2. After trying to swap between controllers and moving the p600 controller from one PCI slot to another, somehow my logical volumes are gone. Do I have any chance to recover the volumes after fixing the connectivity problems?

Thanks

[Ali · Accepted solution]

Hi there,

Dl380 G4 comes with smart array 6i controller and used to connect 6 internal scsi hard drives installed on the server.

P600 is a SAS controller which can not be used to connect SCSI hard drives. so what is purpose of using P600 controller?


Raid information (RIS copy) is stored on the hard drives so other PCI SAS controller movement will not impact logical drive. however, smart array controller may disable the logical drive incase of physical drive movement or internal problem with the logical drive.

things that you can check
1. power off the server (take out the power cables also)
2. take out the P600 controller and put the hard drive to their original location if changed/moved.
3. boot the server from smartstart CD and access the array configuration utility to see the array/logical drive status.
4. If logical drive is detected but has cross mark, right click on it to see if you get enable logical drive option.
5. if no logical drive found and all the hard drives are listed under controller to be configured under a new array then you might have lost the raid information.

if you have lost the logical drive and data is quite critical then you may have to contact some data recovery company as they have utilities/tools which can recover the data even from the failed harddrive

hope this helps,
thanks
Aftab

[DragosPD]

Thanks for your promptly answer.

Yes, you're right, I have a smart array 6i controller not P600 - I didn't know the name by hard and due to a bad search on google I ended up here with a wrong name.

Anyway, the type of the controller is not an issue here. I understand from you that the raid info is stored in scsi drives so a controller replacement will not affect the configuration.

Do you think that this raid info can be lost if the drive is not defect and I didn't confirm any type of erase/overwrite in BIOS and other tools?

What do you suspect that could be the origin of the problem?

Thanks

[Ali]

HI,

smart array 6i is integrated on the system board so you might have another PCI SCSI raid controller. what is the controller model?

if drives are not detected during post or in array configuration utility then there could be various issues.

backplane, controller or cable issue. replacement of these could isolate.

else if you have identical server, you may move these hard drives in the same order to other server and check whether it boots or find logical drive.

thanks,
aftab

[DragosPD]

Once again I messed up with the info :(. Ok, the on board controller is 6i but I have another one on a PCI slot: smart array 6400 which is backed up by batteries.

Here is the complete story:

I buyed the server second hand with four hard drives. I bought two new.

The server is used from time to time (there are days when it is not used at all) but I plan to go with it in production so it should be rock solid ;).

During the normal use, I've noticed that some hard drives are getting their error led amber, even those new. The pattern of the affected drives is generally respected but I'm surprised that the newly two drives are most affected. After a restart, doing something with the cable, backplate hard-drive it seems that problems dissaperead for a period but after that, on a hard usage, re-appers. Now it seems that I lost the logical drives which will put me in pain to restore the entire SO and applications.

Today I will check with another SCSI cable if I can restore the drives, if not - what should I do? Buy a new backplane?

Thanks,
Dragos

[DragosPD]

Right now, by using the 6400 controller I cannot see three hard drives from six. If I change the cable to the 6i controller one of that three failed drives becomes available.

[Ali]

Hi Dragos,

thanks for sharing the detailed information.

Dl380 G4 has two types of cabling.

Simplex configuration
all six drives can be configured on single channel

Duplex configuration
it devides backplane into two
4 hard drive and 2 hard drives
so you can configure two channels or two different controller to create seperate logical drives.

check page no 69 for simplex and duplex configuration in dl380 g4 user guide.
http://www.mc4.army.mil/help/techdocs/DL380G4ServerUserManual.pdf

coming to your question:
Since you have lost everything I would suggest you to check the scsi backplane cabling and make it simplex configuration first.

run the erase utility from smartstart CD to clear everything from the harddrive.

Update dl380 G4 server by running the firmware maintainence CD. (this will update all the firmware on the server)

Run the smartstart CD again and lauch the ACU (array configuration utility) see if you are able to see all the harddrive and able to create array as per your requirement.

if all the hard drive are detected fine, just install through smartstart and setup the server.

later on you may run array diagnostic utility within windows OS to diagnose any controller or hard drive issue.

HP SmartStart CD x32
http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=397634&prodNameId=3288132&swEnvOID=1005&swLang=8&mode=2&taskId=135&swItem=MTX-e796972577c54175a51b8af444

Smart Update Firmware DVD
http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareDescription.jsp?lang=en&cc=us&prodTypeId=15351&prodSeriesId=397634&prodNameId=3288132&swEnvOID=1005&swLang=8&mode=2&taskId=135&swItem=MTX-3380ae888f86436aac54c149eb

at this point no hardware replace is required unless you setup and run diagnostic to verify the issue.

thanks,
aftab

[DragosPD]

Hello, I've just managed to obtain the latest ADU report taken on fifth of december. It is attached here, if someone can help me in order to diagnose the problem.

Thanks

[DragosPD]

I fixed the problem by replacing the backplane. Now I'm struggled with the recovery of the array configuration. When I boot up there is a messsage regarding changing the positions of the hard drives - and it directs me to run ACU in order to find the older positions. I ran ADU but I don't know how to find this information. In ADU,by selecting my array there is no info about my old RAID configuration, it seems that everything is reset. I need advice in order to recover my lost raid configuration, if it is possible...

[Ali]

Hi There,

Please share the ADU report so that we can identify the position and advice you.

else

Boot from the smartstart CD and run Array Configuration Utility (ACU). check the status of logical drive and problem area. it will also guide you if drive possition has been changed.

Also share the screenshot of ACU logical view.

thanks,
Aftab