We had two disk failures on a RAID6 volume on our 3220, in about two days.
Web gui shows volume as degraded; we extracted one drive marked as red in GUI, the other failed disk (also marked as red in gui) was let inserted and powered.
I can't understand led pattern of failed disk still in place, there's no red LED in front of disk tray and blinkin pattern seems as the disk is still Ok and similar to other disk that are GREEN on web gui.
Maybe i miss some of the workings of this NAS model? (i must say i'm quite new to NETGEAR professional products).
The information on page 71 appears incorrect. It talks about the top tray LED as showing activity and the bottom tray LED as showing health "on all trays except RR4360". But pages 43-44 say something different for the RN3220 and RN4220 - the top LED is labeled there as disk power, and the bottom LED is labeld as disk activity.
And the specifics on the bottom LED on page 46 is confusing. It says
Blinking. The disk is active.
Blinking on/off/on/off pattern. The disk failed or is faulty
But it doesn't say clearly how to tell the difference between those states.
I don't own a rackmount model, so I can't clarify this for you. I'm pointing out the issue so someone from Netgear can (and hopefully fix the manual).
Thanks for your reply. I also hope in feedback from Netgear...
Checked with RAIDAR, it reports disk as failed, also.
Sincerely i cannot find different led blinking behaviour (front led, in disk tray) compared to the other disks, disks that both gui and raidar say are ok.
Another thing i must point out, we saw a terrible performance degradation in SMB service from our RN3220, started a couple of days ago, and i cannot connect it at other events, other that 2 disks pulled in failed state. I must say i'm talking about SMB share utilized by VEEAM as backup repository (we adopt compressed, deduplicated veem hyper-v backups).
It's an expected behaviour? Coming from experience only with hardware raid cards on HP server for example, or enterprise storage systems such ad IBM storwize or DS3512, i was used not to have significative perfomance impact on degraded raid volumes.
Another thing i must point out, we saw a terrible performance degradation in SMB service from our RN3220, started a couple of days ago, and i cannot connect it at other events, other that 2 disks pulled in failed state. ...
It's an expected behaviour? Coming from experience only with hardware raid cards on HP server for example, or enterprise storage systems such ad IBM storwize or DS3512, i was used not to have significative perfomance impact on degraded raid volumes.
I haven't run with degraded volumes much. But of course the system does need to reconstruct the blocks it can't read from the parity blocks. So there will significantly more disk I/O, and some delay due to the reconstruction.
Since one of the disks is still in the NAS, the system likely is trying to read it. That might also contribute to the performance issue (since the system won't reconstruct the data until after the read fails).
When replacing drives, you should insert them with power on. Try booting with just the good drives, check that the system shows the empty slots as empty, then hot insert one of the new drives and see if it starts to sync. Using a PC to remove any existing partitions on the new drive before insertion can also sometimes help.
failed disk (in gui) think was the problem; when removed, leaving NAS with 2 disks missing (raid6 volume), then inserting one new disk, first rebuild/resync was quite fast (about 6/7 hours).
Then inserting another new disk (the last missing to complete all the slots), this time rebuild/resync required about 24 hours.
Maybe is wanted behaviour, maybe not, i can't say,
Want to thank evebody helped me, contributing to this post.
