New RN104 owner - concerned about <6.1.5 node size issue

[rotordave]

Hi,

Having read the thread "Readynas 104 drops out when copying to it" (http://www.readynas.com/forum/viewtopic.php?f=65&t=73537), I'm now concerned about which firmware version I had when my data volume was created. I purchased the 104 only last week and have spent a lot of time setting it up and copying data.

I have installed 4 Seagate ST4000DM000 4TB drives and I'm specifically concerned about this problem (in 6.1.4 and prior) that when reaching 70% of capacity used, the system starts to lock up.

During initial setup I let the ReadyNAS create the volume *before* upgrading.

I understand that, among other "improvements", the node/leaf size was increased in 6.1.5 and this resolved the problem.

My questions are:
* How can I determine the node/leaf size of my file system and therefore whether I have the problem?
* How can I determine the firmware version I had prior to upgrading to 6.1.6, and therefore which version was used to create the volume?
* What "workarounds" were identified to eliminate the symptoms? I read about disabling anti-virus, snapshops, etc...

Basically I can't find any bottom-line info about this, surprising since it seems to be a big issue.

Appreciate any info you can provide...


Thanks!

[mdgm-ntgr]

rotordave wrote:
How can I determine the firmware version I had prior to upgrading to 6.1.6, and therefore which version was used to create the volume?

Download the logs, extract the contents of the zip file and look at initrd.log

[rotordave]

Thanks mdgm. So the output of cat /var/log/frontview/initrd.log was

[2014/03/18 08:17:39] Factory default initiated due to new disks (no RAID, no partitions)!
[2014/03/18 08:18:00] Defaulting to X-RAID2 mode, RAID level 5
[2014/03/18 08:18:26] Factory default initiated on ReadyNASOS 6.0.8.
[2014/03/20 00:09:03] Updated from ReadyNASOS [b]6.0.8[/b] to 6.1.6.

This would seem to suggest I am subject to the issues of the old firmware.

Where to from here?

[mdgm-ntgr]

If you maintain a good backup (you should do this regardless) you shouldn't need to do a factory default just yet, but if you encounter major performance issues then you should consider doing one and restoring your data from backup.

[rotordave]

Thanks for your reply.

If possible, I'd still like to know the answer to these questions:
* How can I determine the node/leaf size of my file system and therefore whether I have the problem?
* What "workarounds" were identified to eliminate the symptoms? I read about disabling anti-virus, snapshops, etc...

On the second question - were the workarounds, if there are any, confirmed to address the issue?

I'd rather do the factory reset now than later, given I'm halfway through copying data, but don't want to do it without exhausting all options.


Thanks

[rotordave]

OK, found that the command:

btrfs-show-super /dev/md127

...gives the following info:

nodesize                4096
leafsize                4096

So that confirms the problematic file system, as I understand it should be 32768 with the new firmware. SO frustrating, as I was thinking of doing a factory reset after firmware upgrade given the history with the 4K blocks on the NV+, obviously I didn't learn from that.

All that remains is to understand if there were any workarounds.

[mdgm-ntgr]

With the latest firmware, performance should be O.K. even with the old leafsize, however if you have the time a factory default would definitely be the way to go. The workarounds you mentioned are the key ones.