HP ML350 G4 munching hard drives

[Thomas Mortensen]

Hello,

I have a HP ML350 G4 with a smart array 641 controller and 6 hd's configured in RAID 5.
The server is a domain controller, dhcp, dns and fileserver in a small company. Every 2-6months I have to replace one of the hard drives because of failure, even though the traffic on the server is fairly low. Over the time I have had this server HP support have sent me 6-7 hard drives. Their latest claim is that it is perfectly normal that a HP server destroys a hard drive every 4-6 months..
The hard drives I am using is 72.8GB 10000RPM wide ultra SCSI drives.
Does anyone have the same problem or a solution to end the hard drive munching?

(pardon my bad english.)

[juan quesada]

please attached ADU to diagnose

[juan quesada]

...sorry, attach

[KarloChacon]

hi

uh too much hard drive but I would say:

update firmware for HDD if done already either HDDs backplane (drive cage is faulty) or power backplane is not properly handling the current/voltage that goes to backplane and damaging HDDs

regards

[Thomas Mortensen]

Please find attached ADU report.
I have firmware upgraded all HDD's exept the one I changed yesterday.
HP have changed the whole backplane of the server, but I'm not shure if they did something with the powersupply.
Do you think the powersupply's to weak to run all HDD's?

[KarloChacon]

hi

I am not saying power supplies obviously since they are part of the power subsystem (can affect too)but I meant power module where power supplies connect (Power supply backplane 365065-001) but I think if you got the non hot plug model your power supply will be directly connected to SB and drive cage (HDDs backplane) so there is not power backplane like hot plug models for your server

regards

[Nalin malik]

hello,
yes in hp servers like dl380 g8 is also having same issue.
Actaully iam a RE over at customer side in that place normaly there are 2 or 3 hddd fail in 1 or 2 weeks and sames specs of the hdd
but in this case we can only take the precautions:-
update the hdd firmware
controller firmware
bios update
and check the powercondition that is necessary in point of parts failure
and one last important thing
enviroment condition as well as temperature on the site
can u tell me the eviromental condition of ur location means any type of dust or drain is going through ur location

[Thomas Mortensen]

We have a clean airconditioned server room.
The room temperature is set to 20deg.C
There's only small amount's of dust in the room.

Thanks for the answers :)

[juan quesada]

Hi,

your HDD's have too many timeouts, this is a lack of communication between the controller and the HDD's, because of the fact that the controller cannot sense the drive during its threshold it will mark the drive as bad , replace the backplane and SCSI cables to fix the issue

say thanks with points

Regards,