JorgeB

Yes, I was surprised as well, very little overhead when compared for example with PCIe.

See here for some expander tests, scroll down:

ata-TOSHIBA_HDWD130_37F65K8AS@ ata-TOSHIBA_MQ01ABD100_Z3AJC14TT@ ata-TOSHIBA_MQ01ABD100_Z3AJC14TT-part1@ ata-WDC_WD30EFRX-68AX9N0_WD-WMC1T1300432@ ata-WDC_WD30EFRX-68AX9N0_WD-WMC1T1300432-part1@ ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0425030@ ata-WDC_WD30EZRX-00MMMB0_WD-WMAWZ0425030-part1@ ata-WDC_WD30EZRX-00SPEB0_WD-WCC4E1ZH8492@ ata-WDC_WD30EZRX-00SPEB0_WD-WCC4E1ZH8492-part1@ ata-WDC_WD30EZRX-22D8PB0_WD-WCC4N1YP88NY@ ata-WDC_WD30EZRX-22D8PB0_WD-WCC4N1YP88NY-part1@ ata-WDC_WD60EZRX-00MVLB1_WD-WX31D944CS0X@ ata-WDC_WD60EZRX-00MVLB1_WD-WX31D944CS0X-part1@ nvme-TOSHIBA-RD400_@ nvme-eui.e83a9702000018f5@ nvme-eui.e83a9702000018f5-part1@ usb-Kingston_DataTraveler_2.0_64006A7807FAB031D96709AA-0:0@ usb-Kingston_DataTraveler_2.0_64006A7807FAB031D96709AA-0:0-part1@ wwn-0x10501182279695618049x@ wwn-0x10501182279695618049x-part1@ wwn-0x11233951013753212929x@ wwn-0x11233951013753212929x-part1@ wwn-0x12438954472907100161x@ wwn-0x12438954472907100161x-part1@ wwn-0x16616003131295092736x@ wwn-0x3635306224004976640x@ wwn-0x3635306224004976640x-part1@ wwn-0x6330196881211281409x@ wwn-0x6330196881211281409x-part1@ wwn-0x8622904561878847489x@ wwn-0x8622904561878847489x-part1@

Probably.

Attached. tower1-unassigned.devices-20170716-1445.zip

Yes, I know the NVMe is not identified correctly, still it shouldn't appear as an available unassigned device. Attached. tower1-diagnostics-20170716-1213.zip

Nope.

NVMe cache device is on the unassigned devices list with a strange identifier:

Any dual core Sandy Bridge or newer is enough.

As my sig indicates I use dual parity on all my servers that have 12 or more data disks, although 2 disks failing at the same time is rare, it happened to me on 2 occasions during a rebuild, not a complete disk failure, but a few read errors on a second disk during the rebuild making some of the rebuild files corrupt, at the time dual parity was not yet available. Also feel that checksums are invaluable, not because of bit rot per se, but for when a problem like above happens, or a disk redballing during a write operation, without them there's no way of knowing if or which files are corrupt.

If you want to read: https://www.kernel.org/pub/linux/kernel/people/hpa/raid6.pdf

There's no magic: https://wiki.lime-technology.com/Parity#How_parity_works

Yes

Wrong slash, \\ not //

You can use rsync together with the user scripts plugin.

Supermicro announced the new AMD server boards, I find the H11SSL series particularly interesting, 16 SATA + 8 SAS ports on some models!

I've only seen those spikes with scan users shares enable, and I don't use included or excluded folders.

No, the libvirt.img stores all the VM XMLs.

That only happens if you set "Scan user shares" to enable, default is disable and it should be left like that.

LT just posted their intention to include UD as part of the webGUI, maybe they will also make it support extra pools.

Thanks for adding that, technically the problem is only when partitioning the disk, formatting is OK, but probably best to leave it like that or it may be confusing for some users.

I made the same request earlier basically for the same reasons, so unknowing users don't shoot themselves in the foot, dlandon said he didn't plan to change that and I respect that since he's the one doing the work, but in case you need to help a user it's very easy to partition a GPT disk (any disk >2TB) outside the array, you can use: sgdisk -o -a 64 -n 1:64:0 /dev/sdX This is the same command unRAID uses and the user can then use the UD plugin to format the disk and it can be used as an array disk later. MBR disks (any disk up to 2TB) are different, unRAID generates the MBR contents in a buffer and then writes it to sector 0 of the device, so I don't know a single line command to use, but by now most new disks should be GPT anyway.

Post always the diagnostics zip instead. You CPU is overheating: Jul 5 17:32:57 Tower kernel: CPU7: Package temperature above threshold, cpu clock throttled (total events = 1)

There is only one ST8000DM0004. You can use the speed test scrip/plugin to confirm the rpms, it should start at ~220MB/s, no idea how to confirm the cache but it has to be what's on the Seagate specs.

Cache buffer - 256MB Recording method - Perpendicular Edit: See above