Posts by NoidPurity

    Just an update.


    The virtual disk in Proxmox is VirtIO SCSI with Cache "Write back", if someone needs the info. :)


    I created a "standard" encrypted device from within OMV Web GUI and created a ext4 filesystem on that encrypted disk, and the performance is good, much better than I got in vmware ESXi.

    Hi,


    I'm running OMV virtual on a Proxmox Host.


    I'm using virtual disks on a LVM-Thin and the physical disk is an old WD20EFRX.


    I have done some benchmarking with this command,

    fio --name=random-write --ioengine=posixaio --rw=randwrite --bs=64k --size=256m --numjobs=16 --iodepth=16 --runtime=60 --time_based --end_fsync=1


    Here is the results of the benchmarks,


    VirtIO SCSI no encryption ext4

    WRITE: bw=51.9MiB/s (54.4MB/s), 2993KiB/s-3552KiB/s (3065kB/s-3637kB/s), io=13.0GiB (15.0GB), run=275997-276014msec


    VirtIO Block no encryption ext4

    WRITE: bw=41.1MiB/s (43.1MB/s), 2473KiB/s-2871KiB/s (2532kB/s-2940kB/s), io=11.5GiB (12.4GB), run=287090-287102msec


    VirtIO SCSI no encryption xfs

    WRITE: bw=14.3MiB/s (15.0MB/s), 735KiB/s-1266KiB/s (752kB/s-1296kB/s), io=1978MiB (2074MB), run=131204-138065msec


    VirtIO Block no encryption xfs

    WRITE: bw=10.4MiB/s (10.9MB/s), 588KiB/s-823KiB/s (602kB/s-843kB/s), io=2039MiB (2138MB), run=195132-195315msec


    VirtIO SCSI no encryption btrfs

    WRITE: bw=19.2MiB/s (20.2MB/s), 984KiB/s-1642KiB/s (1008kB/s-1681kB/s), io=1784MiB (1871MB), run=91527-92795msec


    VirtIO Block no encryption btrfs

    WRITE: bw=11.4MiB/s (11.9MB/s), 673KiB/s-774KiB/s (690kB/s-793kB/s), io=8077MiB (8469MB), run=710188-710607msec


    VirtIO SCSI encryption ext4

    WRITE: bw=19.9MiB/s (20.9MB/s), 1227KiB/s-1305KiB/s (1257kB/s-1336kB/s), io=3525MiB (3696MB), run=177168-177181msec


    VirtIO Block encryption ext4

    WRITE: bw=19.4MiB/s (20.4MB/s), 1139KiB/s-1328KiB/s (1167kB/s-1359kB/s), io=3614MiB (3790MB), run=186075-186081msec


    As I understand it should both BTRFS and XFS be faster than EXT4? But it does not look like this in this benchmark. XFS might only be faster on bigger files?


    I also have AES-NI activated for the VM on the CPU, so I tought the encryption should be faster, but it's 30MiB/s slower with encryption enabled.

    What am I doing wrong? 8|


    Thanks.

    Hi,


    I'm using an encrypted storage drive, and after a reboot, when I have to unlock the encrypted volume, I also have to remount the NFS shares under /export.


    Could this be triggered automatic after the encrypted volume is unlocked?


    Like the way FreeNAS is doing, it asks if you want to restart file share services after you unlock an encrypted volume.


    Thanks.

    I have this working with no issues, I have no extra options set, no log queries set, and the only other difference is my dns server is pi-hole.
    Ok this is going to be a 'suck eggs question' on any network machine are you 'releasing' the current ip address and then 'renew' so the machine sends a dhcp request. Anything on my network then appears in the leases tab, and any static entries I have created are entered in the static entries tab.

    Ofc, i do an IP release and renew... but they are timed out. When I then enable the DHCP on my router again and try to renew the IP it works directly... So somehow the DHCP part of dnsmasq isn't starting.... Any ideas or should I ask in the Debian forum? :D

    Did you try it without the extra options? You shouldn’t need dhcp-authoritative and expand-hosts is already written to the config file by default. And just to make sure, your network adapter is actually eth0? Also not sure why you only want it to listen on eth0.

    Yes, I have tried without the extra options also. Well, yes, if I check with ifconfig the adapter for the LAN is eth0. Well, I don't wan't the DHCP to listen on my VPN tun0 for example.

    Hi,


    I can't get the DHCP server to function with dnsmasq and can't use it to lookup local domains. I have not the DHCP-server activted for the moment, because it didn't work, so I use my routers for the moment. But when I active the one in the dnsmasq plugin and disable the one in the router I can't get any IP or information from the DHCP in dnsmasq.



    Bash: /etc/resolv.conf
    # Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8)
    # DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
    nameserver 127.0.0.1
    search vsk

    My plex user got the right home directory, still error tough. Fuzeman says that installing the libraries the plugin needs with pip might solve it, but I asked him to write how to install the libraries in the bug thred on github.


    This is the only plugin I'm using that have "bult in" libraries, so I don't know if other plugins fails. But I think all plugins that have some kind of libraries or programs "built in" that need to be run, will fail because the data disk i noexec.

    You are right. I was thinking that was default location for the plexmediaserver folder before it is moved to data drive. I am working on something..
    I have to look it up...


    I have already fixed the symlink =) I have another error with the plug-in now, but it's posted to the one making the plug-in on github. =)


    Thanks for the help tough. =)