Double size of R6 Gen1 didn’t work

  • 119 Views
  • Last Post 4 weeks ago
philippe hatstadt posted this 4 weeks ago

I use an old iMac with Sierra OS that cannot be upgraded. I replaced all the 6 1Tb drives in my R6 gen 1 configured in RAID5 with 2Tb drives. The way I did it is to change one drive at a time and rebuild the array which is the exact procedure I also follow when a single drive fails. It took a while and it worked and the Promise Utility shows 12Tb array and 10Tb usable. However; macOS even after a reboot still shows the drive capacity as 5Tb and not 10TB. I did not do anything other than what is required to replace individual drives as I read that RAID5 is supposed to automatically resize the drive to the smallest drive times 5. They're all 2Tb.

So is there actually another setup or config that I am supposed to run? If so, would it be destructive as far as data? 

Order By: Standard | Latest | Votes
R P posted this 4 weeks ago

Hi Philippe,

We have a Knowledge Base article that explains how to do this.

This process is best done by those familiar with the terminal and the Promise Utility CLI.

philippe hatstadt posted this 4 weeks ago

This seems quite complex. I have a TimeMachine backup of the drive's data so can I just reformat the entire drive to the 2x size with the Promise Utility and then restore the TimeMachine backup?

R P posted this 4 weeks ago

Hi Phillipe,

For a long time I did not think it was possible to expand an HFS+ volume above the size of the original disk, it took some time to figure it out.

But as far as I know, this is the only way.

Regarding starting over, the simplist way would be to use the Promise Utility and give the drive a quick init. Look in Background Tasks, it's the blue circle at the top right of the Promise Utility. initialization is one option, and once you select it there is a quick option. This is necessary if you want to use the additional space.

Then use Disk Utility to put a new HFS+ filesystem on the drive. It is suggested that you do not put an APFS filesystem on the disk. Forum posters who have done so found that the storage got very slow as the volume got filled.

philippe hatstadt posted this 4 weeks ago

If I do that, am I going to be able to restore the TimeMachine backup seamlessly? I read somewhere that TimeMachine is based on the UUID of the backed-up drive. By building a new file system as you suggest above, will the reformatted drive have a new UUID or the original will be preserved? At this stage, my TimeMachine is the only data available and no redundancy so I would need to be 100% sure I can restore from TM with no risk. 

R P posted this 4 weeks ago

Hi Philippe,

I was not suggesting, I was responding to your mentioning starting over.

I have a TimeMachine backup of the drive's data so can I just reformat the entire drive to the 2x size with the Promise Utility and then restore the TimeMachine backup?

But now that I look a little closer, I'm confused. If, as your last post suggests, the Pegasus is your time machine backup, and you seemed to be suggesting in the quoted text reformatting the Pegasus to use the extra space, are you suggesting that the time machine backup will still be present after reformatting?

The disk would be wiped, you could then re-assign it as a time machine backup target. The initial backup to the Pegasus would take some hours to complete.

If you want to maintain the original backup and use the extra space, the procedure posted is the only way.

philippe hatstadt posted this 4 weeks ago

I am very sorry if I confused you. No the R6 is not my TImeMachine backup, it is my primary external drive for data. I have another standalone non RAID 6Tb external drive that I use for TM. So if I wipe the R6 clean to get to 10Tb and b stead of the current 5Tb, could I easily restore the TM backup on the "new" R6 drive now offering 10Tb of RAiD5 storage? Would the UUID of the build from scratch R6 be the same as the old one? If so, then TM restore should be easy, however long. If not, it's complicated based on a post that I read about restricting to a different UUID. The latter, if trus, seems pretty stupid because most times when someone restores a TM backup would be after a primary drive failure. Whichever new drive purchased to replace it would be unlikely to have the same UUID. So I'm sort of confused by that

R P posted this 4 weeks ago

Hi Philippe,

No the R6 is not my TImeMachine backup, it is my primary external drive for data. I have another standalone non RAID 6Tb external drive that I use for TM.

OK, things make a  lot more sense now. You had not mentioned other storage previously.

Would the UUID of the build from scratch R6 be the same as the old one?

I'm not sure, when I find some time I will test this.

 

R P posted this 4 weeks ago

Hi Philippe,

I created a new Pegasus Volume, saved data about it, deleted and created a new volume and saved data again.

The following things changed.

  • Pegasus Logical Drive Serial Number
  • Pegasus Logical Drive WWN
  • macOS Volume UUID
  • macOS Disk/Partition UUID

If Time Machine uses the Volume UUID, then that will change when creating a new Volume..

philippe hatstadt posted this 4 weeks ago

I decided to try the KB mentioned at the top of this thread as I am not impressed by the issues surrounding the different UUIDs with TimeMachine.

Quick question as it relates to the KB: in the step in the promise utility CLI (i.e. "migrate -a start ..."), the OpStatus is "Synchronizing" and the KB article mentions that this step can be quite long. I don't mind but I have two questions: 

1. Am I going to get a warning when it is done or do I need to periodically do "phydrv" in cli to check the status?

2. On the diskutil cli side, do I need to wait until step 1 above is finished to perform the disk repair command? 

R P posted this 4 weeks ago

Hi Phillipe,

1. Am I going to get a warning when it is done or do I need to periodically do "phydrv" in cli to check the status?

There will be an event. But you can open the Promise Utility and select the Backgound Activities page (blue circle in the upper-right hand of the Promise Utility) and watch the progress in real time.

2. On the diskutil cli side, do I need to wait until step 1 above is finished to perform the disk repair command?

Yes, please do not try and run multiple steps at the same time.  Do the steps sequentially and wait for a step to complete before starting the next one.

philippe hatstadt posted this 4 weeks ago

worked perfectly on OS 13, Thank you!!!!!

Close