Barry Staes

VM and Docker stats (network, cpu, mem) exposed for digestion in any network monitoring system (Observium, LibreNMS, etc) so these do not have to view Unraid as one device. Example via SNMP. Not writing custom graph Prometheus stuff. It is fun, but for this purpose I want something that just works out of the box - just like Unraid.

Thanks, going to try this. And probably also need to adjust “Preserve user defined networks” docker settings. Still a feature to have new dockers start last (instead of first) would be nice to keep things consistent by default.

There must be a better way. I run several dockers from the arr suite and these communicate with each other. For this I use bridge mode, and access them on my LAN via the mapped ports on my Unraid IP. However - for the dockers that must communicate with eachother, this IP apparently does not work. No worries though: So I use the Docker Internal IP in their 172.17.0.x network. All good, until the order of the dockers changes. And the problem is that this network assigns IPs based on the startup order. When the order changes because a new one is added/started before the others, the docker instances can not reach the other instances. * I find it quite inconvenient that when installing Apps, it adds the docker to the top of the start list. It all seems ok, but after a reboot it gets started before the others, and breaks many dockers.

Hi! The Unifi-Poller repository is not using the latest version, and is showing errors (and crashing) with the latest Unifi Network firmware. ``` [INFO] UniFi Poller v2.1.3 Starting Up! PID: 1 ... [ERROR] metric fetch failed: unifi.GetClients(https://1.2.3.4:443): json: cannot unmarshal number 6253.254138136846 into Go struct field Client.data.wired-tx_bytes-r of type int64 ``` Full explanation here https://github.com/unpoller/unpoller/issues/605 CA template appears to use https://hub.docker.com/r/golift/unifi-poller/ with UniFi Poller v2.1.3. The new location mentioned is https://ghcr.io/unpoller/unpoller with UniFi Poller 2.8.1 (currently latest)

Ok thanks, and now i understand what an Unraid root share does. Could not find it in the UI or online. That appears the solution to my problem. PS. I always considered //1.2.3.4/ as a defacto "root share" that exists on all SMB servers. With the way any windows machine browsing would see all shares.

Script sounds great, i am going to try it. I consider using some rsync script like this in user_scripts on Unraid, but before i accidentally use some random old version.. Some questions The initial post (Posted October 18, 2020) contains version 1.5 of the script. Is that the latest version? For external USB you suggest BTRFS , but what about the Unraid Array? Does the filesystem matter if i where to rsync copy from a remote Unraid server to my local Array? And the reverse. I suppose the filesystem must support the links. PS. Some suggestions for the author to help me/people choose and use the script: In the script header, please include a date stamp with the script version. A change log would be even better! In the script header, please include a link to the bespoke source of this script (this thread?) so we can check for a new version. I know i wont remember where i got it otherwise.. Also where can i find the 1.3 if this has problems? Why not give this file a place on github.com to use the versioning there? I'd be happy to help. Their web editor allows to edit the file online, no git expertise or local software needed.

Been considering that as well, it would be the best compromise, by far, so far. @MrGrey not commercial

@theimmortal if you see/reproduce High consistent CPU usage in the Unraid Dashboard again, do this test: Look at the "top" or "htop" command in the console, what kind of CPU usage to you see? As said above i see a 10x deviation between the two.

Similarly my test server with ZFS drives, in the Unraid Dashboard it shows continuous CPU usage (20-60%) when just writing data to the array. With erratic high jumps as well. However .. according to htop it is mostly just 2-5% in total. And htop shows (in this order) krusader (docker) , htop itself , a handfull of shfs , and tailscale (plugin) as highest CPU users. Unraid Version: 6.12.2 Array consists of 2 ZFS drives and one XFS drive. Writing only to a ZFS drive during abovementioned. I did no further tests, just sharing a observation about high Unraid Dashboard CPU graphs. Actual network traffic during all of this is mostly between 40MB .. 250MB per second. (2.5Gbps) unraid2-diagnostics-20230718-2227.zip

I now think its just a bug that this dialog wont allow me to enter "/" or empty. (using Unraid Version: 6.12.2 , Unassigned Devices plugin 2023.07.16) Observations: Works when use "Load Shares", disk mounts OK if the user has read permissions for that share. The "Done" button is not disabled, but it is not working if the "Enter or Select Share" input is an empty value. If i enter "/" here, the dialog is closable using "Done" but the resulting share results in "//1.2.3.4//" (two slash suffix) and wont mount. If i enter " " (a space) here, the dialog is closable using "Done" but the resulting share is "//1.2.3.4/" which looks fine, but still wont mount and shows this error in the log: Jul 16 12:25:10 unraid2 unassigned.devices: Mounting Remote Share '//1.2.3.4/'... Jul 16 12:25:10 unraid2 unassigned.devices: Mount SMB share '//1.2.3.4/' using SMB default protocol. Jul 16 12:25:10 unraid2 unassigned.devices: Mount SMB command: /sbin/mount -t 'cifs' -o rw,noserverino,nounix,iocharset=utf8,file_mode=0777,dir_mode=0777,uid=99,gid=100,credentials='/tmp/unassigned.devices/credentials_1.2.3.4' '//1.2.3.4/' '/mnt/remotes/1.2.3.4_' Jul 16 12:25:10 unraid2 kernel: CIFS: Attempting to mount \\1.2.3.4\ Jul 16 12:25:10 unraid2 kernel: CIFS: VFS: BAD_NETWORK_NAME: \\1.2.3.4\ Jul 16 12:25:10 unraid2 kernel: CIFS: VFS: cifs_mount failed w/return code = -2 Jul 16 12:25:10 unraid2 unassigned.devices: SMB default protocol mount failed: 'mount error(2): No such file or directory Refer to the mount.cifs(8) manual page (e.g. man mount.cifs) and kernel log messages (dmesg) '.

Some automated GUI backup copy solutions i considered. I was surprised there are so many bad options. Veeam (enterprise features, but overkill and expensive) Duplicati (broken, not recommended) Duplicacy (paid yet not improved or maintained, not recommended) Borg + borgmatic (?) Restic (?) Arq5 (?) Urbackup (good, maintained, but different use case) Syncthing (good, maintained, but different use case) Those with a (?) i still need to look at, but seem to have different use cases as well.. are there more i should take a look at?

I would like my old Unraid server to become an offsite online backup for 100% of my server including media. (Yes, i do offline backups too, but only critical personal data. This thread is about offsite online backups!) I am not looking for something more complicated than rsync + a GUI + scheduling/automation options. And if this works, perhaps incremental snapshots. I have looked at some backup softwares (listed in 2nd post below) but lots of these are broken, unmaintained, or not intended for my server-to-server usecase. Rsync is fine but its not a GUI. Doing manual copies in Krusader works for now, but i would rather use something somewhat automated / reproducable. Any options?

If for example i want to add a remote share, and use the GUI, this works fine: It asks a host, user, password, and then makes me pick a share. Remote path would be //1.2.3.4/share123/ in this example, and that works fine. However, it requires a folder suffix. How to add a remote share like //1.2.3.4/ instead? Goal is to backup. I want to copy (backup) all shares of another (Unraid) main-server in one effort to my Unraid backup-server, without having to create a dozen shares, and then just do a copy in Krusader GUI. I have considered (fine) alternatives like rsync before, but i want a GUI. I have considered (awful) alternatives like see here:

New install, first run, something is off with permissions for a new install. See bottom for sus causes and tests. The container was created normally as such: Command execution docker run -d --name='ClamAV' --net='bridge' -e TZ="Europe/Berlin" -e HOST_OS="Unraid" -e HOST_HOSTNAME="bTower" -e HOST_CONTAINERNAME="ClamAV" -e 'USER_ID'='99' -e 'GROUP_ID'='100' -l net.unraid.docker.managed=dockerman -l net.unraid.docker.icon='https://its.ucsc.edu/software/images/clam.png' -v '/mnt/user':'/scan':'ro' -v '/mnt/cache_apps/appdata/clamav':'/var/lib/clamav':'rw' 'tquinnelly/clamav-alpine' -i 51721770e6d4bf942cef86765a8bc9f159e8633e06f90b96327349aab8eee4a0 The command finished successfully! Mind the ID's used. But when started there are errors: ERROR: Can't create freshclam.dat in /var/lib/clamav ERROR: Failed to save freshclam.dat! WARNING: Failed to create a new freshclam.dat! ERROR: initialize: libfreshclam init failed. ERROR: Initialization error! ERROR: Can't create freshclam.dat in /var/lib/clamav ERROR: Failed to save freshclam.dat! WARNING: Failed to create a new freshclam.dat! ERROR: initialize: libfreshclam init failed. ERROR: Initialization error! 2023-03-01T08:54:30+00:00 ClamAV process starting Updating ClamAV scan DB Hint: The database directory must be writable for UID 100 or GID 101 An error occurred (freshclam returned with exit code '2') 2023-03-01T08:54:51+00:00 ClamAV process starting Updating ClamAV scan DB Hint: The database directory must be writable for UID 100 or GID 101 An error occurred (freshclam returned with exit code '2') ** Press ANY KEY to close this window ** Trying to find a cause.. Maybe because my Unraid gave the docker different User and Group ID's ? Maybe because the new `/appdata/clamav` folder was created with less permissions (drwxr-xr-x) than all other folders (drwxrwxrwx) ? => I am running the latest CA Plugin version 2023.02.25 and wil run a test with other dockers soon to compare results.

Thanks for these wonderful Unraid Plugins! I found some links are broken or missing, and for your convenience i tried to make a nice summary for CA User Scripts : Unraid page /Settings/Userscripts > bottom of the page > link "For help with this plugin, visit HERE" => points to https://unraid.net/forum/index.php?topic=49992.0 and shows an empty webpage saying "File not found.". Unraid page /Settings/Userscripts > top-right "Credits" popup, link "Plugin Support Thread" => points to https://unraid.net/topic/48286-plugin-ca-user-scripts and shows a webpage saying "error 404". Unraid page /Plugins > plugin entry "User Scripts" > does not show a "Support Thread" link in the overview, like most other plugins show. Unraid page /Apps (CA) > search "user scripts" plugin > button "Info" > in popup button "Support Forum" => points to https://unraid.net/forum/index.php?topic=49992.0 and shows an empty webpage saying "File not found.". I found no link or mention of the Unraid Plugin User Scripts source repository anywhere in the CA, plugin itself, changelog, this plugin forum thread, or your (Squid's) profile.