Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[PhillB]

Hardware Version: B1
Firmware Version: 1.08
Software Version (Easy Search): None
Harddrive 1: Seagate ST31000340AS 9QJ2DL81 1000 G
Harddrive 2: Seagate ST31000340AS 9QJ2DL8Q 1000 G
Raid 1 Array

Problem Type: Front led does not show when drive has failed

Problem Description: Initially DNS-323 GUI showed Sync Time Remaining:  Degraded, with the drive still showing.  Later the drive no longer showed up at all.  At all times the front panel led remained blue, never changing to amber to indicate a drive problem.

Function Tested: If dead drive was swapped to the other bay in the DNS-323 on powering up both leds still remained blue, so the problem doesn't seem to be slot related.  I'm not sure if this is a 1.08 problem as never had a drive fail under 1.06 or 1.07.  Seagate replaced the drive under warranty and Raid 1 rebuilt fine and showing as Time Remaining: Completed.

Test Procedure (steps to reproduce):   Place defective drive in the DNS-323 and see if the led remains blue or changes to amber as I believe is the expected functionality.

Thanks,

Phill.

[ECF]

If the RAID status is degraded it does not mean that you have a failed drive necessarily so you still may have a blue LED for that drive. If the drives resync and the RAID is now completed your drive is good.

[fordem]

Just a passing comment ...

Test Procedure (steps to reproduce):   Place defective drive in the DNS-323 and see if the led remains blue or changes to amber as I believe is the expected functionality.

This test as described here will always fail in that the LED will not light amber.

If you take a DNS-323 in perfect working order with a single disk drive and insert a defective drive in the remaining bay, the unit will do one of three things, depending on the nature of the defect...

1 - If the unit cannot detect the drive, the LED will not light, it will stay dark
2 - If the unit detects the drive and the drive comes to a ready state, the LED will light blue
3 - If the unit detects the drive and the drive does not come to a ready state, the LED will flash blue and the unit may not complete POST.

For the unit to display an amber failure indication the unit must have the drive installed and functional and then sense that drive as having failed, if a new drive is introduced whether defective or not, the unit will not display a defective indication.

[PhillB]

If the RAID status is degraded it does not mean that you have a failed drive necessarily so you still may have a blue LED for that drive. If the drives resync and the RAID is now completed your drive is good.

Thanks for this.  The drive definately failed, but the led remained steady blue.  Resync could not complete automatically and manually resyncing made the drive dissapear from the DNS GUI all together, though the led remained steady blue for both drives.  Subsequent testing of the drive connected directly to my computer failed to show up even running Seagates diagnostic tests.

[PhillB]

Just a passing comment ...

This test as described here will always fail in that the LED will not light amber..........

Thanks for explaining this, I hadn't realized that was the expected functionality of the led.

Oddly though, even when the drive was put in the second slot the led still became steady blue, but the drive was not shown at all in the DNS GUI.  Maybe a fault occured with the drive that enabled it to come up but nothing further.

[fordem]

Phil

I can't tell you what the expected functionality of the LED is, only what I have observed.

When I first got my DNS-323 (maybe two years ago), it was my first linux based NAS and I was unsure of what to expect and so I spent many hours poking and prodding before committing my data to it.  One of things I was curious about is how it would detect and report a failed disk (which is a critical function to my mind), and so I removed working disks and installed known defective disks in an effort to simulate failures.

[PhillB]

Fordem,

Thanks again for this.  As you say reporting a failed disk is a critical function, which is why I was disappointed that mine failed to indicate it.  At the time I couldn't get e-mail notification to work either so the first I new was when I happened to notice the array was degraded.

This might well have been a one off odd situation though as what made me suspicious of the NAS was that backups to it were failing with CRC errors.  Investigating showed the array as degraded and rebuilding it failed.  With the defective drive removed backups returned to working fine.  It wasn't my understanding of how a Raid1 NAS would work.  I assumed it would continue to function correctly even if there was a single disk problem, but in my case the data integrity failed until the defective drive was removed.

In a way I wish I'd kept the defective drive to be able to do more tests with it, but once it seemed to completely fail I returned it to Seagate to have it replaced under warranty.

Thanks again,

Phill.