HP server: Raid 5 - predictive failure

Try update the firmware and latest BIOS.

Before you update the firmware and/or bios of any kind with a questionable drive system, I would look at the backplane (where the drives get connected to the system) and your raid array controller. I would shut the system down, pull the drives out one by one or mark them so you can keep them in order, blow some air in the drive cage, reseat them. Take the array controller out and reseat it. Better yet, if you can get a replacment raid array controller of the same model, replace it. Then power the system back up and see what you up against. If the array starts to rebuild let it finish before you do anything else. Once you have a healthy raid, then do any firmware / bios updates. If you update the firmware on your raid array controller, BE SURE to also update the drivers in Windows before booting up with the new firmware. Otherwise, your system may not boot up.

Hard Read Errors:  28536 is unacceptable, the other two aren't so bad. But if that is the replacement drive then it doesn't make much sense since it should have started off at a count of zero unless it was replaced with an equally faulty drive. DF072A9844 doesn't come up as having any firmware apart from initial release HPD0 but strangely it's only listed under Integrity servers, not Proliants and then only on one single document on HP's site. It's got 20 months service hours on it, so it isn't new. Maybe it's the stats for the old one that you've posted.

If you search the link below then you'll see that that drive model isn't listed at all, whereas if you just search it for DF072A you'll find 6 other drives that must be pretty similar and they are at HPD7 or above firmware.

I would question where this replacement disk came from, I suspect it may be a model that HP has pulled or forgotten existed, there just aren't enough references of that model number on HP's website to give me confidence in it.

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&cc=us&prodNameId=3182562&prodTypeId=329290&prodSeriesId=1157688&swLang=8&taskId=135&swEnvOID=1005 lists all the drive firmware for Proliants and that disk isn't there.

The hard disk was ordered from HP. and I already updated the firmware and driver... didn't help

Now I just noticed that all models of other hard disk are DF072A8B56, but the one one in Bay3 is DF072A9844...... different model hard disk can cause the problem??
1.JPG
2.JPG
3.JPG
4.JPG
5.JPG

Not so much the different part number but that that particular disk doesn't have firmware available for it. I suspect HP have sent you one they've "tested" rather than a new one. They could have just made a mistake by fixing it but forgetting to clear down the S.M.A.R.T data back to zero. Phone them up first thing and tell them they've sent you a pup.

I will try to order another new hard disk and try it again. and post the result here.

Hi I have repalced the hard drive (hard disk 3) and this time It shows passed in array configuration utility and system management hompage. But in HP insight Diagnostics, the new hard disk now shows passed, but the hard disk 4 and 5 still shows error. Is the error normal? Can I ignore the error?
1.JPG
2.JPG

It seems the errors are pointing toward an expired warranty which is not something to worry about. However, it is also reporting errors beyond threshold. Where are these drives coming from? Are these used? If they are saying its beyond warranty I can't imagine they are coming direct from HP. You will need to may be replace these other drives as well but one at a time giving time for the array to rebuild fully.

4 and 5 will show as bad until replaced as it says the read and write hard error count is above threshold.

so are you guys saying the 2 hard disks are going to fail soon?