DeeGee

Somethings changed. Last night, I powered down to add some more disks. Powered back up and checked they appeared OK. I could see SMART at this point on disks that were not in the array. I *think* this was version 6.5, but can't be sure. It was the old branding anyway. Then updated to 6.6.6 via the GUI, rebooted / rechecked the new disks and SMART was no longer available on the same disks that had been working fine 20 minutes earlier. SMART settings until I started looking today were all default, out of the box. Nothing else changed, apart from the update to 6.6.6.

But it worked fine prior to upgrading to 6.6.6...

Ive done some more digging and this is an unRaid bug, not unassigned devices... To reproduce: Set default SMART controller type to "HP cciss" to match the controller. From the Dashboard, click on the SMART green thumb icon of a device that is in the array, click Attributes and you can set the controller port index (screenshot, disk 10) However if you click on the SMART green thumb icon of a device that is NOT in the array, click Attributes and its not possible to set the controller port index and will get a smartctl error (screenshot, sdh) Therefore the bug is that its not possible to see the SMART data on a device not in the array when using HP cciss controller. The impact of this is that if a drive is added to the server, its not possible to check SMART is OK until after it's added to the array.

Ok, so if I change the default SMART controller type in settings to HP cciss, then on each disk in the array, set the disk index to match the HBA port number, I get SMART info out that matches the disk installed. Guess something has changed as this used to work out the box without the override; but nevermind, its easy to fix for disks in the array... However, if a disk isn't in the array and I'm looking at the details on Unassigned Devices, there is no "smart settings" section to set the disk index on the controller, so SMART fails: =======> valid arguments areata, scsi, nvme[,NSID], sat[,auto][,N][+TYPE], usbcypress[,X], usbjmicron[,p][,x][,N], usbprolific, usbsunplus, intelliprop,N[+TYPE], marvell, areca,N/E, 3ware,N, hpt,L/M/N, megaraid,N, aacraid,H,L,ID, cciss,N, auto, test <======= Is the page at url /Main/New?name=sdd part of base unRaid, or part of the Unassigned Devices plugin?

Set as default out of the box. Worked fine on previous versions (I think I was on 6.5 previously) - only stopped working after upgrade to 6.6.6 last night. Assume I need to make a change here... Ill give it a go later and update...

I can probably add something like that for when Supermicro or Asrock are not detected. My preference for auto detecting was to make sure the right board was selected and thus the right ipmi commands. There are also different commands for Asrock boards based on number of CPU sockets. If things aren't working right (usually supermicro due to threshold settings) some may just start changing settings randomly which would send the wrong commands. Granted this would probably be benign but you never know. I'll see what I can do but I also want to keep the webgui as simple as possible. Fair enough - perhaps the best way then would be to have board info in the config file, which can then be edited by hand if its incorrect, along with appropriate disclaimers! Initial value could be set at install time...

Thanks for the confirmation. Thinking about this some more, this affects any host using remote ipmi - you can't assume that the board the plugin is running on is the same as the remote host. Perhaps the best way forward is to have a dropdown on the remote config section that will allow manual setting of the remote host type?

Hi, I have an X11SSH running unRaid under ESXi. There is a LSI HBA and USB card passed through, and all works fine. I have installed the IPMI plugin from the Community Apps section and can see all temperatures / fan speeds fine, but don't have a Fan control tab. Connection is via network. I know fan control is possible via remote IPMI as I can adjust it manually using ipmitool on a Centos VM... I suspect the board is being detected incorrectly; when I run ipmifan --debug, I get a message: Your Intel Corporation motherboard is not supported or setup yet Have I missed anything or does this not work under ESXi, even though it uses communication over IP? Cheers

I have a working dev unraid VM already setup (using plopkexec) so I'll see if I can get something working - PM sent to @Zeron

just had a look on hub.docker.com and it seems there is already a docker image for unifi video. see: https://hub.docker.com/r/pducharme/unifi-video-controller/ should be able to deploy the docker image directly.

I'm not at home right now, but try this (from memory, so backup your XML first ) you need to unmount the disk in unassigned-devices plugin first. <disk type='block' device='disk'> <source dev='/dev/disk/by-id/ WDC_WD20EFRX_68AX9N0_WD_WMC1T1957452' /> <target dev='hdd' bus='virtio' /> </disk>

To be honest, its not clear what you are trying to achieve here. Can you explain? From your original post, you say you want to pass a 2TB drive through to the VM, but the XML suggests you are trying to use a image file as a disk for your VM and that image file is what resides on the 2tb disk; these are two different concepts... Are you trying to setup a UniFi controller? If so, you could always just use the docker version; works great when installed from the community apps plugin...

my first thoughts would be to check there are no usb power-saving options enabled on your bios that would cause the device to disconnect.

Looks to be presented as device hdd on the VM; see if you can see it by typing this on the command line of the VM: fdisk -l /dev/hdd