siege801

Hello everyone, Can anyone please give some guidance on what steps I need to take with this server and array? The disk in question is a 12Tb Seagate IronWolf disk drive. There are three installed, two of which are parity drives, the third is this problematic drive. It's my plan to slowly swap out the 2Tb drives with more of these (or larger if price-point dictates). Diagnostics attached. Many thanks in advance. lucindraid-diagnostics-20230926-1409.zip

Hey Everyone, It's an exciting time - I'm about to make some fairly significant upgrades to my unRAID box. Before I begin, I'd like to run the scenario by you all to get some advice on the procedure. Current Rig: Onboard Intel NIC 6x 2Tb SATA drives connected to motherboard Parity Style: 5x data and 1x parity Array Notes: Array is encrypted 2x 500Gb NVMe drives in use as a striped cache Final Goal: Onboard Intel NIC (management) Daughterboard Dual Broadcomm NIC (some kind of teaming/bonding) 5x 2Tb SATA drives (existing drives minus one) 3x 12Tb SATA drives (brand new) Parity Style: 6x data (5x2Tb, 1x12Tb) and 2x 12Tb parity. 8-port LSI RAID card (brand new) Break the cache striping and continue with one NVMe as cache (single profile) and the other to be used for playing around with the iSCSI integration. Here is the procedure I've concocted, but I'm very open to correction or advice: Phase 1 - MOAR DATA Pre-step: Run a full parity check. 1/ Install the LSI Card and connect the existing 6x2Tb array - confirm all is OK The rest should just be a matter of stepping through the Replacing a Data Drive guide, correct?. 2/ Connect one 12Tb drive and pre-clear it. 3/ Remove the existing parity drive and replace with the pre-cleared 12Tb from Step 2. 4/ Allow the parity drive to be built up. 5/ Add the second 12Tb drive and pre-clear it 6/ Pick out a 2Tb data drive that has seen the most reads/writes and replace with the pre-cleared drive from Step 5. 7/ Allow the new data drive to be built up. (How do I know when that's complete?) 8/ Repeat steps 5-7 for the third new 12Tb drive. Situation Report By this point I should have 8 drives connected to the LSI and one leftover 2Tb drive. I should now have moved from ~10Tb usable to ~22Tb usable Phase 2 - BREAK THE CACHE According to this thread, I need to: 1/ Stop the array 2/ Unassign pool disk to remove 3/ Spin up the array and wait for the device delete to complete 4/ Once the Stop button is available again, I'll know that the rebuild is done Situation Report I should now still have a 500Gb cache, but made up of only one drive I should now have a spare 500Gb NVMe ready to be assigned? I guess I might need to pre-clear it or something first. Phase 3 - UNRELATED / ADD DUAL NIC While I've got the beast in pieces, I might as well install the dual Broadcomm NIC that's been sitting on the tower for a while. The plan here is to play with teaming/bonding configurations and reduce some network bottleneck. As I've mentioned - this is what I've come up with through reading the various wiki pages (and Reddit posts). I am very open to correction or trout-slapping as required. Many thanks in advance!

Yup, now that I've RE-read the definitions of the Share methods, that makes sense, and the goal of Highwater (spinning up the disks as little as possible) makes sense too. Thanks for that. Looks like I sent myself on a wild goose chase.

Hi, Thanks for all of that. I've adjusted the Tunable - not sure I've easily seen a difference in the upload speed, but again, the bottleneck is most likely the USB speed. Meanwhile, rather than re-uploading the data I'm going to try creating a new, temporary share which should then be accessible at /mnt/user/TEMPSHARE. I'll rsync from /mnt/disk1/SHARE to /mnt/user/TEMPSHARE. Then I'll rsync back to /mnt/user/SHARE. If I understand you correctly, that should upload the files more evenly across the share's underlying array?

Hi All, I'm new to unRAID - but loving it so far! I'm not new to the *nix world, so to me the best way to transfer data from my USB drive to my data array was to mount the USB (console -> mkdir /mnt/usb && mount /dev/sdx1 /mnt/usb); and then rsync the data to /mnt/disk1/SHARE/. This seems to have worked fine, but all the data is, of course, only on one drive. I'm looking into unBalance to 'scatter' the data across the other drives. My issue is, if I understand the unBalance process correctly, I can't seem to scatter the data to the drives in such a way that includes the source drive. Eg): 2Tb Disk0 - parity, so N/A in this context. 2Tb Disk1 - contains 1Tb of data transferred from USB 2Tb Disk2 - blank 2Tb Disk3 - blank 2Tb Disk4 - blank 2Tb Disk5 - blank In unBalance, I select Disk1 as the source, but I then cannot select Disk1 as a target in the scatter process. So my question is, how do I scatter the data correctly? Or should I delete the lot and upload the data differently? I went with this method so that my speed was only limited by the USB transfer, not USB plus network. Please let me know if I've done wrong or what I need to do to share out the data. That said, I've seen Reddit posts stating "why bother?" - just let unRAID do its thing over time. Thoughts? Bonus question - when I transfer data from my next USB - should I be disabling the parity drive temporarily to help with transfer? Thanks in advance!

Hi there, I found your post because I was looking for the same answer. In case anyone else stumbles across this post (or if OP hasn't found the answer). For me, after changing the default filesystem from xfs to xfs-encrypted, I then had to: Open each drive (Main -> Click Disk Name -> Change file system to encrypted). Stop the array (Main -> Array Operation -> Stop On the same screen, there was a new prompt stating that I needed to set an encryption key, so I did. Then when starting the array, each of the drives were listed as unmountable. I ticked the box to format them and away it went. NOTE - in my case, I hadn't yet written any data to the array. If you've already got data on your array, THIS VIDEO from Spaceinvader One looks like the best option. Hopefully that's helpful to someone out there.