An unusual query

  • 35 Views
  • Last Post 3 weeks ago
Eiji Iwakawa posted this 3 weeks ago

I have an unusual query (note I'm a relative novice in IT)

I recently joined a company that has been using a Pegasus R6 as a passthrough dock (JBOD) for their archives! They've literally been using it as a dock for HDDs and swapping out old drives and storing them as an archive. I thought this was a little crazy,  so I've been slowly trying to tidy up their setup to something more sensible. I decided to buy a whole fresh bunch of new drives and set the Pegasus up as a RAID 6 temporary backup for some data (whilst putting aside the JBOD HDDs for now. However, they suddenly approached me and asked to retrieve some old data that should be somewhere in the "archives"...

After doing some detective work, I've come to the conclusion that, although they were using the R6 as this passthrough dock, its first iteration was probably as some kind of RAID array - and it's in this original iteration that the data they want is stored (I found a bunch of HDDs that I believe were the first array).

I'm able to easily check the data on the pass-through drives I found (without using the Pegasus and using a simple dock) but the data I need I think is on the drives that were used when it was RAIDed. I'm not sure whether those drives are definitely a RAID array but I'm assuming they are as they are the ones I can't read individually.

My questions are:

- If I swap out my freshly re-raided drives in the R6 for what I believe to be the old array, will I need to first "warn" or do anything with the pegasus/controller itself to tell it to expect a new array? As mentioned, my current array is RAID6, but I don't know what the previous array was. Is the R6 clever enough to understand that a whole new set of drives are being added with a potentially completely different RAID array set up?

- the drives appear to be numbered in the sequence that they were in the R6, but if there was any error in the order, would there be any major consequences?

- can I restore my current RAID6 array, after I'm done recovering the data, without too much fuss? (I don't intend to be swapping out these drives all the time, I'm just trying to deal with the situation I'm in!)

Thanks!

Order By: Standard | Latest | Votes
R P posted this 3 weeks ago

Hi Eiji,

After doing some detective work, I've come to the conclusion that, although they were using the R6 as this passthrough dock, its first iteration was probably as some kind of RAID array - and it's in this original iteration that the data they want is stored (I found a bunch of HDDs that I believe were the first array).

If the drives are in passthu, macOS (or whatever OS) will see the drives individually. You can format them with Diskutil and put a filesystem on them. The only way to raid such drives would be a softraid through the operating system.

If I swap out my freshly re-raided drives in the R6 for what I believe to be the old array, will I need to first "warn" or do anything with the pegasus/controller itself to tell it to expect a new array? As mentioned, my current array is RAID6, but I don't know what the previous array was. Is the R6 clever enough to understand that a whole new set of drives are being added with a potentially completely different RAID array set up?

When the Pegaus boots it reads the DDF on the drives, no array information is on the Pegasus, all the array information is on the drives. So all you need do is power off the Pegasus (or disconnect the TB cable), swap the drives, then power on the Pegasus (or plug in the TB cable).

the drives appear to be numbered in the sequence that they were in the R6, but if there was any error in the order, would there be any major consequences?

It should be OK if the drives are out of order.

can I restore my current RAID6 array, after I'm done recovering the data, without too much fuss? (I don't intend to be swapping out these drives all the time, I'm just trying to deal with the situation I'm in!)

Yes, just swap the drives with your array back on. Follow the same process.

  • Liked by
  • Eiji Iwakawa
Eiji Iwakawa posted this 3 weeks ago

Amazing! exactly what I needed to know - thank you R P!

Close