Page 1 of 1

Seagate hard drives in RAID

Posted: 2015.01.08. 20:16
by QuantumSpark
Hello,
I have a 5-pro license at home, recommended it to my boss at a small private school, and now we have a 5-pro license here as well. In spite of all that, I'm not terribly familiar with the complexities of drives or this utility. I may be including useless info, and leaving out important stuff!

At the school, I've recently been using a portable install to check the hard drives on our servers after some intermittent but potentially worrisome noises. On two servers, we're using Seagate drives in a RAID array (ST1000NM0023 2TB x4 in one and ST3146356SS 500GB x4 in another). Both are setup in such a way that the RAID hosts C and a second partition.

In each of these setups, a short self-test done on one of the Seagate drives gives an initial estimate of 1-2 minutes, but never seems to finish. The "Response" code while in progress is 0xFFFFFFFF. The progress bar seems to complete, but then jumps back to about mid-way and 'sticks' there. Furthermore, I cannot seem to cancel the test; clicking that button merely hides it as if I'd hit Background, and I have to completely exit out of HD Sentinel and restart it to actually cancel.

(Oh - slight addition / wrinkle, I just noticed that one set of Fujitsu drives (MBA3300RC ~500GB x4) configured similarly tests like the Seagates above, while another set of Fujitsu (MAX3036RC 200GB x5) in RAID with three Hitachi drives tests as described below.)

Testing some other drives in our servers also has a somewhat different though still unexpected result; after 30 seconds of the short self-test, it jumps to "Successfully Completed," no matter where on the progress bar it was (including less than half-way). The Response at completion is 0x40.

Thank you for a great program, and thanks (to whomever) for your assistance and advice with this problem.

Re: Seagate hard drives in RAID

Posted: 2015.01.09. 11:30
by hdsentinel
Generally, the issue may be related to the actual disk (RAID) controller where the hard disks are connected.
Some of such controllers (more precisely the drivers) may not provide real, accurate information when the hardware self test started.
This can cause that
- the disk test seem stuck at a point (seem no advance), the test seem not completes, the progress bar seem stuck at center
- the disk test seem complete very quickly
- the disk test can be started, but when Hard Disk Sentinel sends the cancel command, the test may still seems running.

These are not rare issues, especially when drives configured as RAID, as then some RAID controllers may not allow these kind of special tests (which test the drives indepependently in the array, not together as an array).

Generally, in such situations (as described in the Help), it may be better to perform the Disk menu -> Surface test -> Read test, which will perform complete testing of the array.

If you use Report menu -> Send test report to developer option, I can check the actual disk controller, the driver version, the "raw" details provided when Hard Disk Sentinel detects the self test status.
You may even use Report menu -> Send test report to developer option two-three times, eg. in normal situation (when test is not running), some when drive(s) are being tested (and you click on the Background button to let it run) and after you try to cancel the test.
This way I can compare them, verify the reported information and check the possibiliies, for example if it's possible to improve anything (for example by a driver version update).