Unraid OS Plus v 6.12.13 running on HP Z800.

 

Hi folks, I'm in desperate need of help to rectify a situation I stupidly got myself into.

Although I have been using Unraid for a couple of years now, I still consider myself to be a Noob here (also, after this, of somewhat below-average intelligence as well). This description is probably going to be a bit longwinded but I want to be thorough, so please bear with me.

Referring to this screenshot:

 

 

As you see my array is comprised of a 12TB Parity dve and 8 HDDs of varying sizes. The sizes shown in the screengrab are not the actuals in all cases. I will refer to each HDD by both its Linux mount point & the last 4 characters of its s/n for clarity. The actual sizes are as follows:
 
    parity (sdd)- 12TB
    sdf (3EUP)    - 10TB
    sde    (T92Z)    - 10TB
    sdg    (3K90)    -  4TB
    sdi    (RNVX)    -  4TB
    sdh (JLLP)    -  4TB **
    sdb    (43DF)    -  6TB **
    sdj    (DF29)    -  4TB
    sdc    (1B7Z)    -  5TB
    
    I draw your attention to the drives annotated by **. As you can see there are discrepancies. 
    43DF is a 6TB drive. JLLP is actually a 4TB drive. The mismatches in the drive sizes came about because up until now I had to use a mix of SAS2 & SATA ports. The SATA ports have a 2TB limit on what they recognise as the maximum drive size.
    
    SO here's where I screwed the pooch:
    
    Last week DF29 was in the array recognised as a 2.2TB drive. It was sitting at around 1.98TB total usage.
    I installed a newly acquired LSI 2011-8i IT card into the server. Unfortunately, it came with the wrong break-out cables - they were integrated SAS data & power connectors, rather than the solitary SAS data connector. I have subsequently ordered new breakout cables which arrive today.
    Anyway, being smart (I thought) and impatient (I was) I connected another 4TB drive (JLLP) and transferred all the data from DF29 to it because I thought I was going to have to reformat DF29 and reinsert it into the array when I got the new cables.
    
    COMPLICATING Factor - I noticed 4 unconnected SATA ports that I had originally thought were connected to an SAS backplane for the Original Z800 drive caddy. Realising that this did not make any sense, I thought I might have found extra SAS2 ports that would not have the 2 TB limit. I had another 1TB drive that was unused and decided to connect it to see if Unraid recognised it in UD. It did. I thought I was onto something.
    
    So, I then proceeded to exchange DF29 & JLLP. I used the New Config tool to change the drive order and stipulated that Parity had not changed as per a post I found on the boards. I then reconnected the drive bay containing JLLP to one of the "found SAS2 ports" - except I DIDN'T - I actually reconnected the cable from 43DF to the new port. Blonde moment I know - but that's an achievement in itself 'cause I'm totally Bald.
    
    I then restarted the array in Maintenance mode and found that the new port was actually a SATA, not SAS2, and I got the drive wrong. Unraid then told me that when I started the array again 43DF would be overwritten and then I proceeded to utilise my extensive vocabulary of expletives.
    
    I then tried to put everything back to the way it was originally, and this is where things are at now.
    
    Here's my Plea For Help:
    
    Can anyone give me guidance on how to extricate myself from this self-invoked clusterF@#k of a situation??? I am aware that I may end up losing some data (would prefer not to if at all possible) but would like to minimise that - IE I want to protect 43DF & its 6 TB. If I need to lose the data I transferred onto either or both DF29 or JLLP (it's the same) so be it - Cost of being a Dipshit! 
    
    I just reread this all and it makes sense to me, but I am sure you all will have questions. I have attached a copy of the diagnostics data. If you need any more info please let me know.
    
    One last thing - I am a Kiwi down in NZ, and we have a tradition for buying beers for those who help us out of jams like this. Logistical problems aside, happy to do that in this case for the Legend that is going to save my sanity!!
    
    
    Cheers,
    
    GeeTeeNZ
    
    p.s.  We also have a tradition of throwing sarcastic windups at those who have fucked up so please - "Fill ya Boots" - I'm in need of some new material. :-)

xeus-diagnostics-20211207-1525.zip

The diagnostics appear to be from a different server, wrong ones?

Sorry JorgeB, was old copy - correct server tho. Try this one.

xeus-diagnostics-20241007-1127.zip

So if I understand correctly:

 

- old disk6 is assigned to disk5, so you should just need to correct the assignment to fix that that one, i.e., assign the current disk5 as disk6

- old disk5 was being used as 2.2T and is now being detected as 6TB

 

Is this correct?

4 hours ago, JorgeB said:

So if I understand correctly:

 

- old disk6 is assigned to disk5, so you should just need to correct the assignment to fix that that one, i.e., assign the current disk5 as disk6

- old disk5 was being used as 2.2T and is now being detected as 6TB

 

Is this correct?

No, not exactly.  the 2 drives in question JLLP & 43DF (those are the last 4 chars of serial)  in reality are 4TB & 6TB respectively. They were only reported by unraid as 2.2TB because of the old SATA ports on the Z800 I connected them to. 

 

I have spent the last three hours changing the physical positions of the drives, inserting them individually, basically every possible configuration a number of times. I changed the ports they connected to, with no success. What was strange is that the Logical drive position (e.g., Disk 5 or Disk 6) would change depending upon which solitary drive I had inserted. 

 

So, in the end,  I said f@#k it and used the New Config tool again, reset the configuration, and started the array. All drives are now reported in their correct positions and sizes, and Parity is being recomputed. From the quick check I have made all data seems intact. 

 

I don't know how to explain the bizarre behaviour that Unraid has displayed. 

 

In any case, your comment about changing back the drive positions started the recovery process I guess. Thanks for that. I'll put some money on your beer account later in the week when I get paid. 

 

Cheers,

GT