Jump to content

JorgeB

Moderators
  • Posts

    67,499
  • Joined

  • Last visited

  • Days Won

    706

Everything posted by JorgeB

  1. New beta aligns partitions for non rotational devices on the 1MiB boundary, you can try that, it can be considerably faster for some devices, requires wiping current pool though.
  2. You need to format the array device first.
  3. Docker image is corrupt, delete and recreate, also good to check this for better pool monitoring.
  4. It should, wait for the sync to finish and try again.
  5. SMART looks fine, UDMA_CRC errors are usually a bad SATA cable, replace it and check that it stops increasing, also note that any read errors on another disk during a rebuild would result in the the rebuilt disk being corrupt, unless you have dual parity.
  6. There are checksum errors on cache: Jul 26 15:41:37 Unraid kernel: BTRFS error (device nvme0n1p1): bdev /dev/nvme0n1p1 errs: wr 0, rd 0, flush 0, corrupt 3, gen 0 This suggests a hardware issue, you'll need to delete the corrupt files are restore form backups, scrub will identify them all on the syslog. Also your CPU is overheating, check cooling.
  7. That shouldn't happen, unless there are some reads/writes going to another array disk at the same time.
  8. Logs are still being spammed, very difficult to look for anything else.
  9. Nothing jumps out, you could try the latest beta.
  10. There can always be some disk/controller combinations that don't work correctly, e.g., I had issues before with some HGST 2.5" drives and LSI, 2.5" Seagate SMR drives and Asmedia, and another model with JMB, but forget the model.
  11. I didn't say you can't copy small files, but small files will always be much slower, does the pool perform normally with large files? Also try this:
  12. Because it's not user set unfortunately, I did ask to use the auto option for that, and that if reconstruct write is enable it should only use reconstruct write, there's also a report about it, you can add your opinion there, the more people ask the more likely Tom will take a look at it. It will alternate as it detects multiple device activity in the array. Because normal writing mode (read/modify/write) requires that parity be read before any block is written.
  13. Rsync is not built for speed, still that's slow, make sure you're not using compression (-z), also OSX has some known issues with SMB. Like mentioned part of the problem might be OSX, while some performance loss is normal using user shares vs disk shares, I can easily get around 800/900MB/s from Windows to Unraid (or vice versa) to user shares.
  14. Should work, but if it doesn't create the new flash drive manually with the latest release then copy just the config folder form the old one.
  15. When the rebuild finishes run xfs_repair on that disk and post the output.
  16. dd is known to give very bad results with users shares, did you try an actual transfer?
  17. Currently, and against some users wishes, including my own, Unraid disables turbo write when i/o from multiple array disks is detected.
  18. 2nd rebuild will be identical to the 1st one.
  19. Small files are not ideal to test speed, is speed normal with large files? You can also try the new beta, it aligns SSDs on the 1MiB boundary (requires reformatting the pool) and that usually results in better performance. Example of a transfer of 3 large files totaling about 15GB, destinations is a raid5 pool with 5 cheap 120GB SSDs:
  20. There's no way to find that out, if there was an unclean shutdown since last check it was likely because of that, if there wasn't it's not a good sign, and if you get anymore on future checks you should investigate.
  21. No one knows exactly why this helps in some cases, redoing the flash drive might also help.
  22. Diags are just after rebooting so not much to see, try this.
×
×
  • Create New...