This topic has been marked solved and closed to new posts due to inactivity. We hope you'll join the conversation by posting to an open topic or starting a new one.
We still have a large number of old 2100 units at our car dealerships for backup purposes and I did visit one of these dealers today to physically move one of his servers. These servers usually run various backup routines via NFS to the 2100 nightly, so I attempted a login. No go.
I ran the RAIDar discovery and found the unit somehow resetted itself to DHCP and failing to find a server, set itself up to 192.168.168.168.
So I reconfigured the network settings and logged in. No shares, no volumes. Clicked on the drives button lower right and found drives 2 and 3 dead, drives one and four with SMART warnings in the log.
I have the following questions:
Has anyone had a similar experience with a 2100 losing its network configuration and just "resetting" itself?
The drives (that were still alive) had about 66K hours or close to 8 years (WD 500GB Enterprise) in the SMART data. I always paced server drives at about 5 years - is that still correct?
The dealer would be prepared to spend the $$$ on new drives. I could not test the basic operation of the 2100 with the failed drives though. After removing the two dead candidates, the system would still not let me build a new volume as it did not "see" the 2 drives. Does the 2100 only operate with 4 drives installed or should one healthy drive be sufficient to let the system build a volume?
Thank you in advance for your input and best regards from Germany,
The drives (that were still alive) had about 66K hours or close to 8 years (WD 500GB Enterprise) in the SMART data. I always paced server drives at about 5 years - is that still correct?
Backblaze uses consumer drives, and found that 80% were still in service after 4 years. Their projections say that 50% would still be in service after 6 years (and they see a annual failure rate of about 11% after 4). I don't how this compares with enterprise class.
@vmanns wrote:After removing the two dead candidates, the system would still not let me build a new volume as it did not "see" the 2 drives. Does the 2100 only operate with 4 drives installed or should one healthy drive be sufficient to let the system build a volume?
The NAS can create a jbod volume with one drive, and RAID-1 volume with two. If you were trying to create RAID-5 (flexraid), then of course you would need at least three.
@vmanns wrote:Has anyone had a similar experience with a 2100 losing its network configuration and just "resetting" itself?
That is unusual. If someone accidentally did an OS reinstall, then the network config would be reset to DHCP and the admin password would be reset to netgear1.
If the OS partition became full, then the NAS configuration can become corrupted. Possible I guess, and that could explain why the volume can't rebuild.
You could try a factory reset with just the two working disks installed.
The drives (that were still alive) had about 66K hours or close to 8 years (WD 500GB Enterprise) in the SMART data. I always paced server drives at about 5 years - is that still correct?
Backblaze uses consumer drives, and found that 80% were still in service after 4 years. Their projections say that 50% would still be in service after 6 years (and they see a annual failure rate of about 11% after 4). I don't how this compares with enterprise class.
@vmanns wrote:After removing the two dead candidates, the system would still not let me build a new volume as it did not "see" the 2 drives. Does the 2100 only operate with 4 drives installed or should one healthy drive be sufficient to let the system build a volume?
The NAS can create a jbod volume with one drive, and RAID-1 volume with two. If you were trying to create RAID-5 (flexraid), then of course you would need at least three.
@vmanns wrote:Has anyone had a similar experience with a 2100 losing its network configuration and just "resetting" itself?
That is unusual. If someone accidentally did an OS reinstall, then the network config would be reset to DHCP and the admin password would be reset to netgear1.
If the OS partition became full, then the NAS configuration can become corrupted. Possible I guess, and that could explain why the volume can't rebuild.
You could try a factory reset with just the two working disks installed.
One last question: If I interpret the ReadyNas HDD compatability list correctly, the 2100 is officially limited to 3GB drives. Is there an "unofficial" limit for larger drives?
If I interpret the ReadyNas HDD compatability list correctly, the 2100 is officially limited to 3GB drives. Is there an "unofficial" limit for larger drives?
Netgear stopped updating the HCL for the legacy NAS a long time ago, which makes it of limited value. I always recommend going with either NAS-purposed drives (WDC Red, Seagate Ironwolf) or enterprise class. Those should work reliably in any NAS.
OS 4.2.x systems can accept larger drives, but there are some limits related to volume size. The volume size max is 16 TiB; and you cannot expand a volume more than 8 TiB from it's original size. Note that both limits apply to the delivered volume size, not the raw disk capacity. The XRAID volume size formula is "sum the disks and subtract the largest". So 4x5TB will work; you can get a bit higher if you mix disk sizes (3x6TB+5TB for instance).
OS 6 systems don't have these limits, and it is possible to convert your NAS to run OS 6. It is not supported by Netgear, but many people here have done it. The process is destructive, so you'll need to restore the data from backup.
I did read up on the process to upgrade from OS 4.2.x to OS 6 - I failed to locate the necessary files that are mentioned on differing sites (the links are probably too old and therefore broken):
R4toR6_Prep_Addon.bin
and
R4toR6_latest-beta.bin
Do you have an idea where I can obtain these?
There was also frequent mention of a problem with the USB ports after the upgrade? Do you have any insight on this as we do use the USB ports for data transfer...?
I did order 4 x 2TB WD Enterprise drive to replace the failed 500GB units. I did install an older Barracuda 250GB drive to test the NAS - all o.k.
I appear to have painted myself into a corner under the best intentions though. Our servers rely on NFS shares for backup purposes to the NAS. I updated the firmware from 4.2.15 to 4.2.31 to get the unit updated to the last release. Unfortunately this update replaced the NFS protocol with the Apple file protocol (AFP). Is there a way to activate NFS under 4.2.31 somehow? This is REALLY odd.
Thanx again for your help. It is REALLY appreciated.
I did read up on the process to upgrade from OS 4.2.x to OS 6 - I failed to locate the necessary files that are mentioned on differing sites (the links are probably too old and therefore broken):
There was also frequent mention of a problem with the USB ports after the upgrade? Do you have any insight on this as we do use the USB ports for data transfer...?
The ports should work, I think the issue there is that the path isn't always the same when the disks are inserted, and that can interfere with backup jobs. I don't know of any workaround.
