ReadyNAS Pro: Volume size of more than 13,6TB not possible?

[APcMini]

Hi!

I know it's not supported officially, but I bought four Hitachi 4TB 7K400 (HDS724040ALE640) and put them into my ReadyNAS Pro Business. I also got two WD 3TB (WD30EZRX) drives in this NAS. After a factory reset I chose SINGLE redundancy X-RADI2. After initialisation the NAS said the drive in bay 6 is on rebuild, which was the first strange thing. After that, I recognized, that for ALL 6 drives 2790 GB were allocated. Firmware is: RAIDiator 4.2.20-T42

I then changed the drive in bay 6, did another factory reset and got the same problem, waited for the RAID sync to be finished and got the message "Found space that can be used to expand capacity, expansion process will start during next boot." All 4TB drives now were allocated with 3721GB. After the reboot the "Volume expansion failed". Well. I don't really know what to do. It is not crucial for me to get this NAS up and running in the next few days, but I wanted to get above the 16TB-limitation, so I can do online upgrades again. I now have 14.919.168.712.704 bytes usable.

All logs are available under http://apnet.cc/log/* , for example:
http://apnet.cc/log/expansion.log
http://apnet.cc/log/initrd.log
http://apnet.cc/log/mdstat.log
http://apnet.cc/log/partition.log
http://apnet.cc/log/dmesg.log
http://apnet.cc/log/system.log
http://apnet.cc/log/volume.log
http://apnet.cc/log/raid.conf
http://apnet.cc/log/kernel.log

If you think I should try to get two more Hitachi 4TB, it's ok. Btw.: I think, that on my first install the drive in bay 6 also needed to rebuilt.

Thank you so much for your comments and hints in advance!

[APcMini]

Anybody who can tell me why the drive in bay 6 is failing? I think for the problem about the space after a factory default I found the reason: To use the full capacity, all drives have to be same size? Is that correct?

[dsm1212]

It looks like your getting a 3TB stripe and a 1TB stripe which is expected to use all the space. I think this is your problem:

Code: Select all

[2012/03/29 21:38:06  2104] LINE 448: exec command: echo "DOING_F_RESIZE" > /.os_V_E_continue
[2012/03/29 21:38:06  2104] LINE 4249: exec command: resize2fs -pf /dev/c/c
resize2fs 1.42.1 (17-Feb-2012)
resize2fs: New size too large to be expressed in 32 bits

[2012/03/29 21:38:06  2104] Expand second phase error 10: resize2fs, err=0x100
[2012/03/29 21:38:06  2104] File system expand failed


It's having trouble adding the last 3TB to the 3TB stripe. I thought the capacity you are trying to do was supported, but maybe it's coded to do some special layout that is not kicking in for the unsupported 4TB disks. I've got 4TB disks in mine too, but not at this high a capacity so I'd like to understand this too.

steve

[evan]

No, it don't need same capacity drives, you do factory default with 4 TB x 4 and 3 TB x2, it will build 3TB capacity with RAID5 on each drives, you see disk 6 is on rebuild, that is right, it is building RAID 5, system will tell you "Found space that can be used to expand capacity, expansion process will start during next boot' after RAID sync on each HDD''s 3TB capacity, then reboot system, system will do vertical expansion for the rest of 1TB capacity on 4TB HDD.

[dsm1212]

It looks like your getting a 3TB stripe and a 1TB stripe which is expected to use all the space. I think this is your problem:

Code: Select all

[2012/03/29 21:38:06  2104] LINE 448: exec command: echo "DOING_F_RESIZE" > /.os_V_E_continue
[2012/03/29 21:38:06  2104] LINE 4249: exec command: resize2fs -pf /dev/c/c
resize2fs 1.42.1 (17-Feb-2012)
resize2fs: New size too large to be expressed in 32 bits

[2012/03/29 21:38:06  2104] Expand second phase error 10: resize2fs, err=0x100
[2012/03/29 21:38:06  2104] File system expand failed


It's having trouble adding the last 3TB to the 3TB stripe. I thought the capacity you are trying to do was supported, but maybe it's coded to do some special layout that is not kicking in for the unsupported 4TB disks. I've got 4TB disks in mine too, but not at this high a capacity so I'd like to understand this too.

steve

Some generic linux info about this here: http://blog.ronnyegner-consulting.de/20 ... ow-solved/

Seems that resize2fs doesn't support crossing the 16TB boundary. So you need the >16TB filesystem created from the start unless netgear has some other fix that became available recently, but resize2fs is giving the error so it appears not. I'm not sure why the NAS is doing this in two steps instead of one. Maybe a Jedi can explain. Is it possible you've tried two drives in slot 6 that might have issues?

steve

[evan]

Hi dsm1212,
You are right, it has 16TB limitation for expansion, this is EXT4 limitation, you must create > 16TB capacity at first.

[APcMini]

Thank you for your help. I now got two another 4TB Hitachi inside my NAS and it's just working great. Also there was one little confusion again:

After factory default, the drive in bay 6 needed to be rebuilt (with 118MB/s average ) again. What's wrong with bay 6?

Now my NAS is working 2 days and there are only minor RAM-usage-problams, but it works fine and the reading speed is OK, too.

[dsm1212]

You've had 3 disks have a problem in that slot so it might be worth a support call to have netgear look at your system. Especially if the system is new and still under warranty. You could look at /var/log/messages to see if that drive showed any problems during the init which excluded it.

steve

[APcMini]

I tried to locate that log-file, but didn't find any file/folder named messages. Well. The NAS is stable after the initiate setup, so I think I wouldn't touch it, but thank you for your suggestion. I will try that in a while.

[mdgm]

After factory default, the drive in bay 6 needed to be rebuilt (with 118MB/s average ) again. What's wrong with bay 6?

That's normal. The drives have to be synced sector-by-sector to setup the RAID. After a factory default the drive in the last bay is the one that shows as being in resync.

[APcMini]

OK. Then everything is running very smooth, thanks!