Migration of RN426 from X-RAID RAID 5 to FLEX-RAID RAID 6

Thanks. I have the log file ready to send- 1.3MB.

I don't use Google Drive, but can send by WeTransfer if you prefer.

I appreciate your help.

I will hold off making any changes to the 426. My first preference is to not have to go down the route of a Destroy and rebuild, if possible.

Kind regards,

Warren

I'm sorry, but I don't think I explained myself clearly in the prior reply. I am not on any of the social media platforms so I cannot use PM. I'm on email, and you are welcome to email me, if that would help. I can send you the link to WeTransfer (for the log file) or I can place the link to the WeTransfer download on this forum if that would help. I don't think the log files have any confidential data on them?

Thank you again,

Warren

Perhaps I can add that yesterday, as a parallel attack on the problem,  I logged a case for support on this issue. Case no 41388938, in case you are able to access the case notes and the attached files.

Thanks,

Warren

---

WarrenT wrote:

 I am not on any of the social media platforms so I cannot use PM.

---

"PM" -> Private Message.  It's a facility in this forum, so you can use it.  Just click on the envelope icon in the upper right of the forum page.

---

WarrenT wrote:

I don't think the log files have any confidential data on them?

They do, and that is why JohnCM_S asked you to send the link via PM.

Hi Stephen,

Thanks for educating me!!

John,

Apologies for the delay, but I have now done a PM to you which i hope gets through OK.

Thank you both for your patience,

Warren

Sounds like you need to get help from Netgear.  I could give you the steps to do it manually via SSH, but it's ptretty involved when you have two RAID groups.

As far a switching back to XRAID, I don't think you'll be able to.  I swiched to FlexRAID, did nothing that should have made the volume not XRAID compatible, and can no longer switch back to XRAID.  It says I cannot because I have "expanded volumes", even though that expansion was done by XRAID long before I switched.  This limitation is not included in KB22813 , and I'm pretty sure has not always been there (I swear I did it before).

---

Sandshark wrote:

 

As far a switching back to XRAID, I don't think you'll be able to.  I swiched to FlexRAID, did nothing that should have made the volume not XRAID compatible, and can no longer switch back to XRAID.  It says I cannot because I have "expanded volumes", even though that expansion was done by XRAID long before I switched. 

Support might be able to get the system back into XRAID mode though.

Gentlemen, Thanks for your advice about the switch back to XRAID. My 6th disk was hot inserted and then I tried, unsuccessfully, to switch to Flex Raid and add the 6th disk as a Parity disk to move to RAID 6. But the system would not allow this because of 2 RIAD groups. Reading the above comments, today I tried to switch back to X-RAID and was not able to for exactly the reasons set out above. The error massage was exactly the same as above. So now, my 426 is in FlexRaid, Raid 5, and can move neither forwards to RAID 6, nor move back to X-RAID.

The thought crosses my mind as to whether there is a fault in the software?

I just wonder if the only option is to backup the data, deep format all the 6 drives to remove any and all vestiges of the old volume history and rebuild the 426 again as X-RAID, RAID 6??

Thanks, Warren

---

WarrenT wrote:

The thought crosses my mind as to whether there is a fault in the software?

 

---

I think there is at least one: the system should have allowed you to add parity (migrating from RAID-5 to RAID-6).

You should have been able to revert back to XRAID too.  But that one might be harder to fix, since there are some flexraid modes that can't be reverted to XRAID.  At the very least they should warn you when you switch to flexraid that the system won't let you switch back.

---

WarrenT wrote:

 

I just wonder if the only option is to backup the data, deep format all the 6 drives to remove any and all vestiges of the old volume history and rebuild the 426 again as X-RAID, RAID 6??

Thanks, Warren

---

I suggest that you just wait on support (+ JohnCM_S ) - I don't see any urgency, since your array is intact and backed up.  Adding parity can certainly be done from the linux command line.  Support might also have some tricks to get the volume back into XRAID.  Waiting on them would also give them a better analysis that would hopefully lead to a software fix.

But if you want to use the "nuclear" option, you of course can.  I don't know what you mean by "deep format", but whatever you are thinking it isn't necessary.  You'd just do a factory default (either from the web ui or the boot menu) after the backup.  Or you'd destroy the data volume from the web ui and create a new one.  Either way you'd end up with a single RAID group.

Thank you Stephen,

I appreciate your quick respons.

1. I will await advice from Tech Support (Case no 41388938 has been recorded as of 21 June - no response yet) and JohnCM_S.

2. If the collective wisdom is the nuclear option, I will ask Tect Support to assist with that.

3. Deep format = external Windows style full format, not just a quick format. I note your comment about using the factory default. The question I have is will that mechanism destroy enough of the deeply embedded data to ensure that the same problem  will not emerge again. I note that OS6 will not allow Destroy of the single Data volume. Ie if there is only 1 volume on the system. 

Q: If I do a factory default reset with all six disks inserted, will the system reboot under X-RAID, RAID 6 automatically?

Many thanks again,

Warren

---

WarrenT wrote:

I note that OS6 will not allow Destroy of the single Data volume. Ie if there is only 1 volume on the system. 

 

No, that's not correct - it will allow you to destory a single data volume.

---

WarrenT wrote:

 

Q: If I do a factory default reset with all six disks inserted, will the system reboot under X-RAID, RAID 6 automatically?

 

---

It would start up as XRAID, RAID-5.  You'd need to switch to flexraid and destory the volume, and create it again with FlexRAID.  Or do the procedure we discussed earlier - which should work with only one RAID group.

---

WarrenT wrote:

 

3. Deep format = external Windows style full format, not just a quick format.

---

That's not necessary, and actually can get in the way. 

---

WarrenT wrote:

 

The thought crosses my mind as to whether there is a fault in the software?

 

I wouldn't lable it a "fault", I'd call it a shortcoming.  XRAID (at least as implemented in OS6) is "just" an automation of features that Linux MDADM and BTRFS provide.  I can give you the commands to build an identical volume via SSH and commands that are not unique to the ReadyNAS.  But the programmers can only account for so many possibilities in building software that automates those tasks.  I don't fault them for that, but I do fault them for not properly documenting those limitation.  Some of the more basic configurations that are not allowed do seem easy to implement (yours being one of them), but one has to draw a line somewhere.

While you can destroy the only volume, you need to uninsall any apps before you do.  I don't know if there is anything that stops you if you do have installed apps, as I have not tried it.

To switch your NAS to RAID6 (which XRAID would have to figure out and perform), you need to do the following:

1. Partition the new drive identically to the old ones:  One partition each for OS, swap, and each data RAID group.
2. Add new partitions to MDADM RAID groups for the same, changing the data groups to RAID6, and let them sync:  OS, swap, and each data RAID group.
3. I don't think anything needs to be done via BTRFS for this because the volume size is not changing, though a scrub might be a good idea.

Stephen and Snadshark,

Thank you gentlemen for your helpful comments and guidance on matters to consider in moving forwards from the current position.

At this time the 426 is running, I think, under FlexRaid - Raid5. The data is secure and backed up.

I am hopeful that Support will soon provide some options to move forwards safely and I will ask their help to guide me in implementing the chosen route. One of my requirements is to stick with a 'standard' implementation of OS6 to ensure future upgrades of OS6 will continue to work in my case.

What seems a helpful feature would be for OS6 to ask you, at the time of first start up, what Raid approach you wish to follow, and offer a number of guided choices explaining the limitations and possibilities as you make each selection. I have not looked at QNAP to see how they approach this aspect.

I can confirm that I have not installed any Apps on my NAS so that is one complexity that is off the table.

With gratitude,

Warren

Hi Stephen and John,

Allow me to thank you both for your great assistance in this conundrum of a case.

I have given thought to what I should do if Support are not able to fix my problematic RN426. I conclude that in such an event the 'nuclear' option is proably the only way forwards. ie Destroy the volume, factory default reset and restore data from my RN424.

I have set out what I believe are the correct steps to be followed in the event that I am forced to follow the nuclear option.

I would welcome you casting your eyes over this to make sure the path set out looks technically correct.

Thank you in advance for any comments you may care to make.

Warren

Your writeup looks correct.  The NAS will come up in RAID-5 (not RAID-6), so you will need to go through the flexraid steps.  Note that you don't need to wait for the RAID sync to complete - you can execute step 6 (and 7) while that is still happening.  That will save quite a bit of time.

On step 11, you don't want the restore jobs to be scheduled, so you do want to make sure that "enable" isn't checked (enable just enables the schedule).  However, you start the job from the web ui - clicking on the settings wheel for the backup job, and selecting "start".

FWIW, if your router supports address reservation you might consider reserving 192.168.1.60 in the router instead of setting a static IP.  That's a bit more robust (making sure the NAS will get still be reachable even if you get a new router that uses a different IP address range).

Stephen,

Many thanks for the review and comments on 3e, 6 and 7, 11. 

Good idea to reserve the IP address in the router. There are 4 separate MAC's for the RN426, one for each eth port. I need to reserve 192.168.1.60 for eth0. Now that the 4 ports are bonded, it will be interesting to see what the router IP table shows for the RN426.

My thanks again for these suggestions. I hope it will not be necessary to trigger the nuclear option and that Support come up with a fix.

Cheers,

Warren