JorgeB

Yep, that's why I mentioned posting the complete diags, cache filesystem is also corrupt, you need to backup and recreate cache filesystem, note that checksum errors are usually a sign of bad hardware, like bad RAM.

You can never hot-swap a mounted disk, hot swap it to avoid powering down the server, you just remove the disk while the server is on, but never while mounted.

That's about right for default write mode, did you try turbo write? P.S. one of your disks is showing some lower performance on the outer cylinders, is that repeatable?

Seems unlike to me for that to be the problem, but please report back.

SAS2 HBAs usually use breakout cables, SFF-8087 to SATA or SAS.

Yes, that will work, it's the model mentioned on this thread.

That controller is x4, it can work on the GPU slot (if available), otherwise you have a 2 port model which uses a x1 slot, there's also one from Asmedia.

Taking another look it see UD is mounting them, and they are all al mounted: /dev/sdn1 5.5T 5.5T 43G 100% /mnt/disks/External_USB30_DISK04 /dev/sdk1 5.5T 5.5T 23G 100% /mnt/disks/External_USB30_DISK01 /dev/sdl1 5.5T 5.5T 27G 100% /mnt/disks/External_USB30_DISK02 /dev/sdr1 5.5T 5.5T 25G 100% /mnt/disks/External_USB30_DISK03 /dev/sdj1 5.5T 5.5T 29G 100% /mnt/disks/External_USB30_DISK00 But there are a lot of i/o and filesystem errors and apparently the disks disconnected and re-connected with a different letter, try rebooting, but that doesn't look like a reliable way to use those disks.

Make sure "Initial Memory" and "Max Memory" are set to the same value.

This is the serial.

While it went mostly over my head I did ask LT to look at your post on Wednesday, since it might have good info, so hopefully they took a look.

You should start by running iperf to confirm it's not a network problem.

This is unfortunately typical with most multi bay USB enclosures, all devices have the same serial number: Apr 12 16:32:14 Tower unassigned.devices: Disk with serial 'USB3.0_DISK00_20170331000C3', mountpoint 'External_USB30_DISK00' is not set to auto mount and will not be mounted. Apr 12 16:32:14 Tower unassigned.devices: Disk with serial 'USB3.0_DISK02_20170331000C3', mountpoint 'External_USB30_DISK02' is not set to auto mount and will not be mounted. Apr 12 16:32:14 Tower unassigned.devices: Disk with serial 'USB3.0_DISK03_20170331000C3', mountpoint 'External_USB30_DISK03' is not set to auto mount and will not be mounted. Apr 12 16:32:14 Tower unassigned.devices: Disk with serial 'USB3.0_DISK01_20170331000C3', mountpoint 'External_USB30_DISK01' is not set to auto mount and will not be mounted. Apr 12 16:32:14 Tower unassigned.devices: Disk with serial 'USB3.0_DISK04_20170331000C3', mountpoint 'External_USB30_DISK04' is not set to auto mount and will not be mounted. Because of this only one will work with Unraid, since it requires valid serial numbers to keep track of the devices.

There are issues from boot, especially with one of them, try disconnecting sdf (Serial Z1Z8S9J40000C5451WXZ) and post new diags after a boot. These disks can sometimes come with a custom firmware and have issues working outside the enclosure they came with.

That's a single CRC error, you can acknowledge that, not a disk problem. Problem is filesystem corruption on disk1, check filesystem: https://wiki.unraid.net/Check_Disk_Filesystems#Checking_and_fixing_drives_in_the_webGui

Docker image is corrupt, you need to delete and re-create. Next time please post the diagnostics: Tools -> Diagnostics

Please don't use external sites, you can attach the diags to the forum.

There's a problem identifiing the drives: Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Read Capacity(16) failed: Result: hostbyte=0x00 driverbyte=0x08 Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Sense Key : 0x2 [current] [descriptor] Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] ASC=0x4 ASCQ=0x0 Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Read Capacity(10) failed: Result: hostbyte=0x00 driverbyte=0x08 Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Sense Key : 0x2 [current] [descriptor] Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] ASC=0x4 ASCQ=0x0 Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] 0 512-byte logical blocks: (0 B/0 B) Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] 0-byte physical blocks Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Test WP failed, assume Write Enabled Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Asking for cache data failed Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Assuming drive cache: write through Apr 11 15:04:42 Tower kernel: sd 9:0:2:0: [sdd] Spinning up disk... You can (and should) update LSI to latest firmware, but doubt it will help, looks like some issue with the disks, some times SAS devices come with unsupported features, like format protection, unsupported sector size, etc, can't see in these since even SMART reports aren't working.

Was that a long test? If it's memory and by the parity checks it can go hours without errors. Yes, if with all disks connected on a different controller you still get sync errors board/RAM/CPU are likely the problem, if not it's likely the onboard controller.

SMART tests are failing: SMART Extended Self-test Log Version: 1 (1 sectors) Num Test_Description Status Remaining LifeTime(hours) LBA_of_first_error # 1 Short offline Completed: read failure 90% 177 15033509320 # 2 Short offline Completed: read failure 90% 177 15033509320 # 3 Extended offline Completed: read failure 90% 177 15033509320 # 4 Extended offline Completed: read failure 90% 177 15033509320 # 5 Short offline Completed: read failure 90% 177 15033509320 # 6 Short offline Completed: read failure 90% 177 15033509320 # 7 Short offline Completed: read failure 90% 177 15033509320 # 8 Extended offline Completed: read failure 90% 177 15033509320 # 9 Short offline Completed: read failure 90% 177 15033509320 #10 Extended offline Completed: read failure 90% 177 15033509320 #11 Short offline Completed: read failure 90% 177 15033509320 #12 Extended offline Completed: read failure 90% 177 15033509320 #13 Extended offline Completed: read failure 90% 176 15033509320 Disk need to be replaced.

That's correct, IF the transfer source is outside the array.

Google TB vs TiB

Both checks should be done without rebooting.

RAID controllers are not recommended, and you'll run into trouble if you need to replace it with a different controller, but SMART might be available if you choose the correct default SMART controller type on Settings -> Disk Settings -> Global SMART Settings, try HP cciss.

Array to array copies with Unraid were never fast, mostly because of parity being a dedicated drive, with earlier releases it was a little faster with Turbo Write, still because one disk has to read and write at the same time it was never that fast, but you could get 70/80MB/s, now, and since the write mode auto changes to r/m/w it can never go above 50/60MB/s, so what you are experiencing is normal, and speed will be the same with or without turbo write, turbo write is much faster for any transfers originating outside the array, like your desktop or from cache or any UD device.