DrivePool getting confused about drives

[spuniun]

So I was attempting to migrate my media server to DrivePool from a Storage Space on Win2012 Essentials R2 for the added features it offers.

Sadly, I ran into a pretty severe problem that resulted in the loss of lots of media (not impossible to replace, just time consuming). After messing with it for about a week I determined that DP was essentially getting confused about which drive was which in the case of 2 of 7 devices, both in the same JBOD enclosure connected to the same multiplied port of a single SATA controller.

It appears that during duplication to these drives, DP would mix them up entirely and write content to the blocks of one thinking it was the other and vice versa. This resulted in NTFS filesystem and file corruption (some pretty cool stuff where one video file would basically have the content from a completely different video file inserted into it).

Oddly, of the 7 drives I have, 6 of them are using port multiplying and only these two were getting confused by DP; they are not identical drive models (both are 2TB, one a WD and one Toshiba) and they do not have collided signatures (I even cleared the GPT on each and rewrote them to ensure uniqueness).

But no matter what I tried, if one or both of these drives were added to the pool, DP would invariably corrupt the filesystem when writing to them.

I'm currently migrating what's left of my usable data to FlexRAID and have submitted logs to StableBit for their review, but posting here in case anyone has some suggestion or insight.

[Christopher (Drashna)]

Since you've posted this in a ticket already, I'm not going to try to answer it here, because there is a lot more data there. 

 

However, the entire issue is definitely a very odd one, as you've described it. And hopefully, we can get it sorted out.

[spuniun]

thanks for replying here too, was reaching out to users to see if anyone had encountered similar behavior so no need for official support here

[Christopher (Drashna)]

Oh, I definitely understand that. But from a management/customer service standpoint, it's less than ideal for us. Which is why I replied to it with what I did. I wanted to make sure that anyone seeing it knew that the issue had been addressed.

 

Also, if/when the issue gets sorted out, I will try to make sure the solution is posted here as well (if you don't beat me to it). Worst case, I try to remember every issue I encounter (be it mine or a customers) and use that to help troubleshoot issues in the future.