Migrating from X-RAID2 single redundancy to Dual redundancy

Thank you @mdgm.

Here's mdstat.log:

Personalities : [raid0] [raid1] [raid10] [raid6] [raid5] [raid4] 
md2 : active raid5 sda5[0] sdd5[4] sdc5[2] sdb5[1]
      5846379888 blocks super 1.2 level 5, 16k chunk, algorithm 2 [4/4] [UUUU]
      
md1 : active raid1 sda2[0] sdd2[3] sdc2[2] sdb2[1]
      524276 blocks super 1.2 [4/4] [UUUU]
      
md0 : active raid1 sda1[0] sdd1[4] sdc1[2] sdb1[1]
      4194292 blocks super 1.2 [4/4] [UUUU]
      
unused devices: <none>

And here's initrd.log:

[2010/05/25 06:22:12] Factory default initiated by Frontview!
[2010/05/25 06:42:35] Updated from RAIDiator  to 4.2.11.
[2010/07/09 17:59:10] Updated from RAIDiator 4.2.11 to 4.2.12.
[2010/09/05 03:55:21] Updated from RAIDiator 4.2.12 to 4.2.13.
[2010/09/28 01:19:30] Updated from RAIDiator 4.2.13 to 4.2.14-T11.
[2010/10/12 01:19:28] Updated from RAIDiator 4.2.14-T11 to 4.2.14-T14.
[2010/10/16 19:27:21] Updated from RAIDiator 4.2.14-T14 to 4.2.15.
[2011/04/10 07:49:56] Updated from RAIDiator 4.2.15 to 4.2.16.
[2011/06/01 19:45:32] Updated from RAIDiator 4.2.16 to 4.2.17.
[2011/09/12 08:12:33] Updated from RAIDiator 4.2.17 to 4.2.19.
[2012/05/09 04:24:27] Updated from RAIDiator 4.2.19 to 4.2.20.
[2012/07/03 02:58:58] Updated from RAIDiator 4.2.20 to 4.2.21.
[2012/09/28 18:33:00] Updated from RAIDiator 4.2.21 to 4.2.22.
[2013/07/01 17:38:07] Updated from RAIDiator 4.2.22 to 4.2.23.
[2013/07/27 16:47:48] Updated from RAIDiator 4.2.23 to 4.2.24.
[2013/12/01 01:44:26] Updated from RAIDiator 4.2.24 to 4.2.25.

Since you have updated to 4.2.25 a factory default would give you a clean setup on that firmware.

A factory default would give you 4k sector partition alignment for all your disks regardless of whether they were added before/after you updated to 4.2.12.

When you updated to 4.2.16 the partitioning of your disks was converted from MBR to GPT. With a factory default on 4.2.16 or later your disks are partitioned using GPT from the start.

There may well be some filesystem improvements in the firmware since 4.2.11 which can only be obtained by creating a new filesystem (i.e. if using X-RAID2 when you do a factory default).

Sounds like a factory default on this system would be worth it in the long run.

I'm less concerned about the data since it will be fully backed-up then I am some system tweaks I'd like to preserve. A while ago, I complied and installed a custom build of ffmpeg with the HE-AAC codec for Subsonic and didn't write down all the steps… unless there's a way to save it, guess I'll have to figure out how to build it all over again! :rofl:

Re: new drives - the 4TB drive I'm looking at is the Seagate ST4000DM000. I realize this is not on the HCL for the Pro Pioneer but it is for the new ReadyNAS products and comes highly recommended from a very knowledgeable colleague. Is the HCL for legacy products still being updated? Any word on this particular drive from legacy users? I did a search of the forums and didn't see anything definitive but it seemed good to go.

Get a ST4000NV000 instead. At the moment it seems to be the same price in the US; it is intended/designed for home NAS; and it has a longer warranty than the DM.

Thanks @StephenB, will go that route.

BTW, here's the Facebook post and a link from my colleague regarding enterprise-class drives. It seems Google did a huge test and uses consumer drives rather than enterprise drives for their servers. Food for thought…



http://static.googleusercontent.com/ext ... ilures.pdf

Note the publication date is 2/2007. The study is still interesting, but it would be useful if Google (or someone else) updated it to cover modern drives.

Mr. Schmidt's thoughts on drive temperature are not consistent with the paper he cites. Also he misstates their comments on SMART. They found that after after the first reallocated sector, drives had a 21x greater chance of failure within the next 60 days (compared to drives that had no such failures). This is not not about watching the drive for the first 60 days only. More generally, they only found 4 SMART parameters are "strong", and found that 36% of the drives failed with no SMART errors being reported prior to failure.

Google also clearly does not view all drive models as being equally reliable. "...Failure rates are known to be highly correlated with drive models, manufacturers, vintages..." It would be truly useful if Google (or other folks running large data centers) published their actual reliability data on the drives they use (identifying %failures by model).

Thanks for the clarification!

If you don't mind, I'd like to share your analysis with Adam and that FB thread. Agree it would be a great service if Google made reports like this more public. They can test and collect data at a scale way beyond any other company.

No problem sharing these thoughts.

I have no data on his main point (which is that enterprise-class drives are a waste of money). I've never used enterprise drives in my various NAS, but I did switch to the NAS-purposed drives when they became available. They are close in price to consumer drives, and have better warranties. I guess its fair to say that I suspect he's right (since I've chosen not to purchase them).

While it would be great for the rest of us if Google did report detailed failure data, it could create issues with their suppliers... So I suspect we won't see it.

One more question -

Now that I'm backed up, what's the correct sequence for full factory-default/disk replacement?

Since I'm completely replacing the disk set, can I simply shut down the ReadyNAS, replace all disks, then power on, set up as new and restore data? And wouldn't this allow me to archive the old disk set?

In other words, no need to factory default with the existing disks in place since they're all being pulled.