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.
This is probably a really bad idea, but I am just exploring ideas.
Let's start with info on my current environment:
RN214: X-RAID with 2x8 TB + 2x3 TB
I have copied ~ 6 TB of DVDs to this X-RAID volume.
Total space used on the volume is about 8 TB, and everything except for the DVD copies is backed up.
What I want to do:
Since none of the DVD copies are at real risk of loss, they seem to be wasting a lot of protected RAID storage, so I am considering reconfiguring the NAS to Flex-RAID with a single 8 TB drive as RAID 1 (for the DVD copies), and 3x3 TB drives as RAID 5 for stuff I really want protected.
The problem is that I do not really have a spare 6 TB of storage sitting around that I can use to backup all of the DVD copies. And although the DVD copies are not really at any risk of loss, to re-copy all of those is a LOT of work! So I am trying to imagine the easiest way to make these changes to my NAS.
Here is my idea:
If I remove one of the existing 8 TB drives from the current X-RAID volume, that volume will continue to operate in a degraded status. So I would re-format that now external 8 TB drive to connect as a normal Windows USB share, then back up the degraded share to this USB drive. In fact, there would be enough space on this single 8 TB USB drive to do new full backups of ALL the existing X-RAID shares (but the whole X-RAID volume would remain degraded during this time).
After I have the new backups on the USB drive, I would do a factory reset on the RN214 and configure it as Flex-RAID with the 1x8 TB drive RAID 1, and 3x3 TB drives RAID 5.
Finally, I would restore the backups from the 8 TB USB drive, sending the DVD copies to the new RAID 1 volume, and all the other shares to the new RAID 5 volume.
Question:
Are there any problems with this plan (other than the increased risk of data loss during the time that the existing X-RAID volume is in degraded mode)?
Thanx for any thoughts or ideas you may have on this...
Since none of the DVD copies are at real risk of loss, they seem to be wasting a lot of protected RAID storage,
6 TB would be a lot of DVDs, so I am thinking that re-ripping them would be a very big job. Something to consider in your overall strategy of not backing them up.
If I remove one of the existing 8 TB drives from the current X-RAID volume, that volume will continue to operate in a degraded status. So I would re-format that now external 8 TB drive to connect as a normal Windows USB share, then back up the degraded share to this USB drive. In fact, there would be enough space on this single 8 TB USB drive to do new full backups of ALL the existing X-RAID shares (but the whole X-RAID volume would remain degraded during this time).
After I have the new backups on the USB drive, I would do a factory reset on the RN214 and configure it as Flex-RAID with the 1x8 TB drive RAID 1, and 3x3 TB drives RAID 5.
Finally, I would restore the backups from the 8 TB USB drive, sending the DVD copies to the new RAID 1 volume, and all the other shares to the new RAID 5 volume.
Question:
Are there any problems with this plan (other than the increased risk of data loss during the time that the existing X-RAID volume is in degraded mode)?
This will work, but I suggest starting with just the 3x3TB disks installed. Then after the sync, you'd switch to flexraid and add the 8 TB drive as a jbod volume.
The reasoning here is that the apps and home folders are both placed on the initial volume. It's a bit better to make that redundant.
Since none of the DVD copies are at real risk of loss, they seem to be wasting a lot of protected RAID storage,
6 TB would be a lot of DVDs, so I am thinking that re-ripping them would be a very big job. Something to consider in your overall strategy of not backing them up.
If I remove one of the existing 8 TB drives from the current X-RAID volume, that volume will continue to operate in a degraded status. So I would re-format that now external 8 TB drive to connect as a normal Windows USB share, then back up the degraded share to this USB drive. In fact, there would be enough space on this single 8 TB USB drive to do new full backups of ALL the existing X-RAID shares (but the whole X-RAID volume would remain degraded during this time).
After I have the new backups on the USB drive, I would do a factory reset on the RN214 and configure it as Flex-RAID with the 1x8 TB drive RAID 1, and 3x3 TB drives RAID 5.
Finally, I would restore the backups from the 8 TB USB drive, sending the DVD copies to the new RAID 1 volume, and all the other shares to the new RAID 5 volume.
Question:
Are there any problems with this plan (other than the increased risk of data loss during the time that the existing X-RAID volume is in degraded mode)?
This will work, but I suggest starting with just the 3x3TB disks installed. Then after the sync, you'd switch to flexraid and add the 8 TB drive as a jbod volume.
The reasoning here is that the apps and home folders are both placed on the initial volume. It's a bit better to make that redundant.
so I am considering reconfiguring the NAS to Flex-RAID with a single 8 TB drive as RAID 1 (for the DVD copies),
This would be jbod, not RAID-1. RAID-1 is two mirrored disks.
6 TB would be a lot of DVDs, so I am thinking that re-ripping them would be a very big job. Something to consider in your overall strategy of not backing them up.
This will work, but I suggest starting with just the 3x3TB disks installed. Then after the sync, you'd switch to flexraid and add the 8 TB drive as a jbod volume.
The reasoning here is that the apps and home folders are both placed on the initial volume. It's a bit better to make that redundant.
Thanx for the response. Of course a single disk is not RAID 1 - I knew that was wrong when I typed it, but my brain chose just that moment to remind me how stupid it can make me look whenever it wants! Oh well.
And I appreciate the suggestion on doing the RAID 5 volume first - that was new information for me. I guess I'll get another 3 TB drive and head down this path. The nice side-benefit of this change is that it will also give me back a spare 8 TB drive to use for whatever, and I'll end up with the about the same total amount of NAS storage that I have now (just that half of it won't be RAID protected).
So far so good. I did pull one of the 8 TB drives from the X-RAID volume and use it for the full NAS backup as a USB3 drive, formatted as EXT4. The only slightly odd thing I ran into was that I could not get the NAS to recognize the USB3 drive UNTIL I formatted it for Windows! After that, it immedieately recognized the drive and added it as a USB share, and I was able to use the Admin Page System Overview window to reformat it as EXT4.
Next, I pulled the other 8 TB drive, and with just 3x3 TB drives installed I reset it to Factory Default from the Admin Page. This seemed to work without any problems. Restoring the backup of my Configuration Archive worked perfectly to reset all the old Shares, email, security, etc. I was pleasantly surprised that I did not have to re-do any of this somewhat tedious setup.
But now here is my NEW question:
Even though the \Data volume had no data on it at all (just the empty Shares defined), the NAS has been doing a long, protracted volume Resync, just as if I had done a volume expansion instead of a factory reset. After about 7 hours, it is only about 25% done. The log does not show anythingwrong, just this as the very first message after the reset: "Volume: Resyncing started for Volume data."
It is not that I think there is anything wrong, just that this resync surprised me. Does anyone know why it would need to do this on a brand new X-RAID volume?
just this as the very first message after the reset: "Volume: Resyncing started for Volume data."
It is not that I think there is anything wrong, just that this resync surprised me. Does anyone know why it would need to do this on a brand new X-RAID volume?
This is normal. RAID creates a virtual disk from the physical ones, and the file system runs on top of that virtual disk.
The parity blocks in the RAID array need to be created when the array is formed, and the sync process is the same when the volume is empty as it is when the volume is full.
Not a question; just another interesting (to me) observation on this whole process:
After everything seemed done and stable with the new 3x3 TB RAID5 volume, I added the single 8 TB drive back to the box and configured it as a new JBOD volume. When I did this, I did not first wipe the drive or format it on the ReadyNAS - assumed none of this was necessary because the last use of this drive was in the same NAS, but as part of the now-gone old RAID5 volume that I got rid of.
When I first added the drive to the NAS running Flex-RAID, the Admin Page showed that it still saw the old volume allocations that existed when I first pulled the drive and did the factory reset (there was one big partition marked as a Mirror, and a smaller partition marked as a RAID5). I destroyed both of these old volume remnants in the Admin Page Volume tab, and then made the new volume on the same drive. All of this happened without error. The new JBOD volume was healthy and did NOT show any messages about resync like the new RAID5 volume had done. After letting it sit for a short while, there seemed to be absolutely no activity going on, so I started the full 6 TB restore job from the USB3 drive.
Now here is the new interesting observation: It seems as though there really was a bunch of background overhead going on for this new JBOD volume.
I base this assumption simply on how slowly the restore went for the first 12 hours or so. My original backup of the entire NAS to that USB3 drive moved along around 100 MB/s, and the other restore jobs I had done from the same drive back to the new RAID5 volume seemed to run up in that range too. But this new big restore of all the video files from the USB3 drive to the new JBOD volume was poking along at <17 MB/s! This slow speed was pretty steady for the first 12 hours of the restore job, and then the reported speed began to steadily climb, eventually getting over 50MB/s by the end of the 27 hour job. Since that reported speed was an average over the entire run of the job, it had to be running back up around 100 MB/s during the last half to overcome that slow first 12 hours.
I only post that information back here in case it ever helps anyone else contemplating a similar change to their own NAS. Although I did not run into any problems with this whole process, I would have spent less time worrying and making notes if I had known more about what to expect.
Jbod volumes don't need to be synced, so there is something else going on.
What disk model is the 8 TB drive?
The 8 TB drive is a WD white label WD80EMAZ shucked from a WD Elements case. This is the exact same model as the other 8 TB that I just pulled from the NAS to use as a USB backup drive (and that drive did not experience the same huge slowdown during the full-vlume backup that I ran to it).
WD does not provide any specs on these white-label drives; many people have speculated that they are simply re-labeled Reds, others have found some evidence that they may be Ultrastar drives. While it is possible that the drive is SMR garbage, that seems quite unlikely since WD was only sneaking the secret SMR drives into the Red line for 6 TB and smaller models. And besides, even if the first 12 hours of very slow data transfer was somehow related specifically to the drive model, that would not leave room to explain why the next 15 hours of the same job (all moving identical vidio files) would have run so much faster.
I am wondering if somehow this mysterious slowdown was related to the fact that I did not format or initialize that 8 TB JBOD volume - I just deleted the previous volume segments on the drive by using the Admin Page DESTROY frunction. Since I do not have a clue what might be going on under the covers in the NAS OS, it could be anything.
It is likely it is an SMR drive, which cannot sustain continuous writes without significant slow-down. That can then cause a backlog of ReadyNAS housekeeping functions, which further slows things down, possibly on the whole NAS.
