[05:47] hello all [05:48] I'm getting error msg on moving a directory. mv this /to/path is giving me ----> mv: inter-device move failed: ‘141552’ to ‘/nfs-share/app/assets/attachment/1573414’; unable to remove target: Is a directory [20:54] is ubuntu 18.04 set up for ssh log in [20:54] ? [21:02] emOne: only if you installed openssh server [21:02] oh [21:03] both the live-server installer and and alternative server installer (also mini.iso, i think) offer this option during the installation. [21:03] E: Unable to locate package openssh [21:03] openssh-server [21:04] thanks tomreyn [21:04] (maybe the (default) live-server installer did not offer installing openssh-server in version 18.04.2 and earlier, though) [21:04] tomreyn: I am running ubuntu 18.04 through a proxmox VM image [21:05] you're welcome, emOne. now install your ssh authentication (public) keys and disable password based authentication. [21:05] and move ssh to a different port number than 22. [21:05] going to run a quick update and upgrade first before installing openssh [21:05] tomreyn: what is a good port? [21:06] or does one not speak about that in a public IRC forum? [21:06] 22 ;) [21:07] lol hello tds [21:07] tds: how do you know I am here! [21:08] /nick emTwo [21:08] I just happened to be here anyway :) [21:08] * emOne is emTwo now. I don't know who this em0ne is that you are talking about [21:09] hehe [21:10] emOne: any other non standard port i'd say. [21:10] 222 2222 22222 or be more creative [21:11] I guess that is a good enough method to keep automated scans from discovering me [21:12] the idea there is just to not have those keep your ports busy [21:12] err openssh-server [21:12] it's not really a security measure [21:13] ahh ok [21:13] nssh-server is already the newest version (1:7.6p1-4ubuntu0.3) [21:16] I am going to ask tds over at proxmox [21:17] the openssh server must be running [21:17] it is asking for my pass afterall [21:17] tomreyn: thanks for all the help! [21:18] you're welcome [21:35] emOne: password based root login is not a good idea. [21:36] I am going to change it to public / private key in a minute [21:36] I am happy I got ubuntu booting through proxmox in the first place [21:37] tomreyn: btw is it just me or is the debian SSH console prettier? [21:37] I get colours when I boot into debian [21:37] no colors in ubuntu ssh [21:39] emOne: those can be enabled. [21:40] https://askubuntu.com/questions/16336/how-to-get-coloured-terminal-over-ssh [21:53] hello,a question,my raid controller refuse to show me serial of disk so i cant identify disk when they go bad,i ended up pulling each hard drive and then mark the missing disk with corresponding mountpoint so when a mountpoint fails i sort of know which disk it is that way also i made a id file for mount point in each hard drive root folder so i can properly remount same hard drives to same mount points as some of my program depends on it,my [21:53] this be done any other way? [21:54] im using my raid controller in jbod mode [21:56] Greyztar: depends on the raid controller. most either allow querying / configuration from the running ubuntu system using vendor specific software and / or provide such information from an OOB / remote management system [21:58] tomreyn: my controller is an ancient lsi controller im using tw-cli to manage it but seems very limited,it shows me model name but no serial,its no good as ive got many disk of same size so cant differentiate if im using my other sata controller i can use smartctl which properly shows the serial so can id the disk that way [21:59] same size and make that is [22:05] Greyztar: if the controller / management software doesn't disclose it then you're out of luck, unless it's maybe presented in the bios. [22:05] like in an option rom [22:06] oh smartctl does show it? ok. then hdparm -I should, too [22:10] tomreyn: thanks for input ill try hdparm aswell =) [22:31] Greyztar: how were you attempting to get smart data out from the controller? [22:31] you may need to poke smartctl with some 3ware-specific arguments iirc [22:32] tds: dang it i didnt think of that! Of course that might work cheers! [22:33] Greyztar: take a look at the -d TYPE... section of the smartctl man page :) [22:33] tds: ill try it tomorrow though im soon signing off thanks for the heads up i used it to pull smart info already but in an automated way to email me and such just didnt think i could just do it manually also on terminal of course [22:34] usually the serial number is used by udev to create a /dev/disk/by-id/ symlink [22:34] ohh wait no that was maybe something else hmm ill look into it tomorrow non the less [22:36] TJ-: thanks for that input aswell ill note it all and go at it tomorrow seems weird that the raid controller didnt show it in own software though [22:38] Greyztar: so tw-cli /cX show didn't list the serials? [22:38] TJ-: im afraid no just model number :/ [22:39] Greyztar: was there a "serial" in the header row though? [22:40] TJ-: so frustrating as all searches on internet shows serial in tw-cli interface then again im running ancient hardware maybe i should invest in newer controller,hmm im not sure what mean by header row though theres a model number and some other number after but i couldnt correlate it to the serial number on the disk tohugh [22:42] TJ-: im fairly certain its nothing to do with serial number as its same number on many disk of same size and make [22:42] Greyztar: "/cX show" or "/cX show drivestatus" outputs a table with a header-row, e.g. "Port Status Unit Size Blocks Serial" [22:42] Greyztar: I'm wondering if the tooling you have is an older version that doesn't report Serial, or if you get that but only see the drive model listed there [22:44] TJ-: yeah i think that might be it also,theres no serial head row only model i can see [22:48] Greyztar: there are some interesting observations in this https://serverfault.com/questions/683464/get-disk-serial-number-by-smartctl-under-raid0-of-lsi [22:50] Greyztar: is this a SAS controller? If so, sas2ircu should work [22:53] TJ-: interesting that article indeed,i think its a sas controller yes im a little tired so ill have to end here ,ill report back tomorrow if i got it working again thanks for helping [22:58] if you're using tw_cli, I think you'll end up wanting -d 3ware,x rather than -d sat+megaraid,x, but give it a poke and find out :) [23:50] after months and months of a passthrough win kvm/qemu, an update? (didn't reboot for a week), now stops gpu(nv) to bepicked up by vfio, unlike before. I added softdep nv/nouv* to ...-load.d/vfio.conf or ratrrher in modprobe.d , and it is agaain caught. However, now when I start the vm, which 'starts' the screen is blank and possibly the passed through controller with trhe hids (kb/m), doesn't seem necessarily to switch over (not sure due to [23:51] scree nbeing all blank). recent qemu updates? apparmour? libvirt? something something? where could I find maintainers of these packages? [23:51] [01:34:27] has passthrough issues soared slightly in past week or so when I didn't reeboot.. or maybe fortnight? :) [23:53] wondering if people using server for kvm/qemu machines with windows might have noticed issues with the vfio after an update of say libvirtd/qemu/apparmour/kernel/initrd or similar in past fortnight or so. [23:55] windows is hosting? [23:55] no , of course not [23:55] ubuntu [23:56] windows is a guest machine to do sketchup, archicad , things like this [23:56] any any other programs in need for gpu, the odd game, and any old fashioned progs etc :) [23:56] I had for many months, in modprobe.d : blacklist nouveau and similar [23:56] oh, youre using passthru [23:57] and now I noticed yesterday, that when the card wasn't grabbed it was loaded with nouveau, etc.. so I don't think they [23:57] yes [23:57] I have windows server and win10 running, but use all virtual devices [23:58] for a short year o r whatever, and some updates (I htink once before) made some changes so I had to fix, but didn't take too long, but right now, I get aagain an issue and notice withoout any change in configurations, that during boot it would ignore the blacklist and still bind to nouveau.. making me feel4 it was almost like a compiled kernel module. [23:58] and no virtio [23:58] you mean you pass the devices trhough or? [23:58] I use a mix of passthrough and virtio for devices [23:58] the windows runs on a nvme [23:59] as windows is such a horrible OS for anything computational, I give it as fast stuff as it can have just to make sure it can do simple things.