VRx

The latest image (b41) ships PostgreSQL client 18, so the Catalog backup now also works with PostgreSQL 17 and 18 servers; just update the container, no other changes are needed.

Hi @isvein, sorry for the very late reply. The Tailscale hook fails because this image doesn’t install extra packages at runtime. Its Alpine package repositories are intentionally empty, so the hook can’t add ca-certificates, jq and wget, and the container starts without Tailscale. The DOCKER_MODS suggestion won’t help either: nothing in this image reads that variable (it is handled by the startup system of linuxserver.io images), and the mod would need the same package repositories anyway. For a LAN DNS server, a per-container Tailscale connection doesn’t add much anyway. The DNS records point to addresses on your home network, so a remote device can resolve the names but still can’t reach those hosts. For remote access, put the Unraid host itself on your VPN instead: Tailscale on the host as a subnet router, or the built-in WireGuard VPN Manager. Your devices then reach the whole LAN, and this DNS server along with it.

Native Zabbix agent 7.0 LTS for Unraid. It runs on the host, so it sees CPU, memory, disks, filesystems, network and the array directly, with no container and no bind mounts. Install: Apps → search “Zabbix Agent” (in progress), or Plugins → Install Plugin: https://github.com/vrx-666/unraid-zabbix-agent/releases/latest/download/zabbix-agent.plg How it worksThe plugin downloads the official static agent from cdn.zabbix.com and checks its SHA256. Nothing is rebuilt. Agent 7.0 works with Zabbix server 7.0, 7.2, 7.4 and later. Settings → Network Services → Zabbix Agent: set your Zabbix server, active checks server, hostname and host metadata (for autoregistration). Any other agent option goes under Advanced, one per line. The options are the standard ones from the Zabbix documentation. Apply checks the settings with the agent before saving. If they are wrong, nothing is saved and you see the error. On the Zabbix server, use the “Linux by Zabbix agent” template, plus the “Unraid by Zabbix agent” template for the array and disks (below). SecurityThe agent runs as root (needed for SMART). To keep a compromised Zabbix server from using it against Unraid, only the keys of the Linux template and the Unraid keys are allowed. system.run, reading arbitrary files, web requests and the like are denied. To allow more, add AllowKey= lines under Advanced. The static agent has no TLS, so traffic is not encrypted. Keep it on a trusted network. Unraid templateFor the array and disks there is a Zabbix template: zbx_unraid_by_zabbix_agent.yaml. Import it in Data collection → Templates → Import (Zabbix 7.0 or newer) and link it to the host next to the Linux template. Array state, with a trigger when the array is not started. Every array and pool disk is discovered automatically: status, errors, temperature and SMART health, each with a trigger. Disk data is read from Unraid, so no disk is woken up. A sleeping disk simply has no temperature until it spins up. Temperature threshold: macro {$UNRAID.DISK.TEMP.MAX} (default 50 °C). Set it per disk with a context, e.g. {$UNRAID.DISK.TEMP.MAX:"cache"}. The keys behind it: unraid.array.state, unraid.disks (JSON, read without waking disks), unraid.smart.health[sdX]. Source, details and issues: https://github.com/vrx-666/unraid-zabbix-agent

Support thread for the Bacula File Daemon plugin Application: Bacula 15 file daemon - https://www.bacula.org GitHub: https://github.com/vrx-666/unraid-bacula-fd Plugin URL: https://github.com/vrx-666/unraid-bacula-fd/releases/latest/download/bacula-fd.plg What it isA native Bacula file daemon (bacula-fd) for Unraid. It lets an existing Bacula Director back up the Unraid server itself - the flash drive (/boot/config), VM definitions (/etc/libvirt), appdata, shares - and restore files to their original locations. You need a Bacula 15 Director and Storage Daemon somewhere on your network, for example the Bacula-Server container (also in CA). Why a plugin and not a Docker containerA file daemon has to see the files it backs up. In a container every path must be bind-mounted, and restores land inside the container's view of the filesystem. As a plugin the daemon runs on the host: it sees the whole system without mounts, restores to the original paths, and keeps ownership, permissions, ACLs and extended attributes. Requirements- Unraid 7.0.0 or newer - Bacula 15 Director + Storage Daemon SetupInstall the plugin. The daemon starts and listens on port 9102. Settings > Network Services > Bacula File Daemon: set the Director name to match your Director, Apply, Restart. A random password is generated on first install. The config is checked before saving - if it has errors, nothing is saved and the error is shown. On the Director, add a Client resource with your server's address, port 9102 and the same password. To keep ACLs and extended attributes, add aclsupport = yes and xattrsupport = yes to the FileSet Options. Good to knowStopping the array restarts the daemon. A daemon reading files under /mnt would keep the shares busy and the array would never stop, so the plugin stops it when the array starts stopping (a running job fails) and starts it again once the array is stopped - so the flash drive can still be backed up. A job run while the array is stopped can back up an empty directory and still report "Backup OK". Schedule jobs for when the array is up and check the number of files. Updating the plugin restarts the daemon (a running job fails). The config lives in /boot/config/plugins/bacula-fd/ and survives updates and removal.

Thanks — good to know we see it the same way. A quick update since my earlier post: the image has been rebuilt and now ships Bacula 15 together with Bacularis 6.3.0, which is the current upstream release. I have also reworked how the image is built and how I follow upstream releases, so updates should no longer lag behind the way they used to. Same place if you want to take a look: https://github.com/vrx-666/bacula-server

I don't know if it makes sense. This is a hobby project developed in my free time. Bacularis delivers the application in Docker, including the entire Bacula stack, including the Postgresql database engine. https://bacularis.app/doc/brief/installation.html#install-using-docker You don't need to run the database in a separate container. I don't think this is a good solution, as containers for different applications use their own database servers, which creates a significant resource overhead for multiple applications. I also believe that Docker wasn't designed to package entire application stacks into a single container. This makes diagnostics and maintenance more difficult, but that's just my opinion, and I know that many users find this convenient. Docker file is here https://github.com/vrx-666/bacula-server/tree/master if You want to play...

@pk1057 thx for info, I will check entrypoint

@pk1057 try to backup configuration files (move to different place on disk), and try to run container without any configuration files. If it go correctly You could stop Container, and move configuration files back on place, should work without problems (but one thing: user/password/name/host for database should be the same) btw. there is posibility to restore database from bacula backup, but this is more difficult, and can take some time

@murkus I don't know if it is stiil relevant but I found some important info in bacula docs: 1. Maximum Part Size This directive allows specification of the maximum size for each part of any volume written by the current device. Smaller part sizes will reduce restore costs, but will cause additional but small overhead to handle multiple parts. The maximum number of parts permitted per Cloud Volume is 524,288. The maximum size of any given part is approximately 17.5TB. 2. it is useful to set the Maximum File Size to a value smaller than or equal to the configured Maximum Part Size BUT You have to remember that Bacula is not small app, and is was not designed to work inside container, this is my small project to backup content of virtual machines on homelab. Maybe it will never work inside container like it could on physical server.

You can do anything, could be "/mnt/cache2" inside docker container. Next you need to add same Path to monitoring (zabbix webui)

Here is docs which I used with testing configuration. https://docs.baculasystems.com/BEDedicatedBackupSolutions/StorageBackend/cloud/cloud-backup.html#cloud-accounts You must do like a ceph endpoint in the exapmle. I've found somewere that for other cloud storage than amazon You must use: UriStyle = Path Default is Virtualhost.

Yes it is different, I can find manual if You need, I found it previous for testing, but I'm not using cloud backup. I must test it one more time. I thought I had already improved it

What data? Zabbix is monitoring system which needs minimum three parts: zabbix-server - to put it simply: it collects data zabbix-webinterface - is responsible for the ability to display data in the browser zabbix-agent - must be installed on the host from which it collects monitoring data and sends it to zabbix-server (active or passive) it depends on what data you want to monitor, templates are related to the operating system, application or other parameters that can be collected and monitored

Today's image have amazon cloud driver added.

@murkus There is a bug while compilation Amazon Cloud Driver plugin building Bacula from src. I've made an issue for Bacula developers.

@murkus thanks for feedback. Yes, the cloud/s3 driver is not working yet Due to the version change and potentially reported problems, I wanted to stabilize this version. I would like to launch the cloud/s3 driver as soon as possible, but at the moment it is difficult to promise a specific date.

@trashit You must set "cache disks" path were You have mounted cache on Unraid, for example if You have /mnt/cache1234 You must set /mnt/cache1234 Container path can remain unchanged, but You must set at Zabbix Web what You want to monitor as container path, not Unraid path. Monitoring physical disks is tricki, You could mount /proc/diskstats to the container as new "virtual" path, and then prepare Zabbix to monitor diskstats from mounted file instead default /proc/diskstats (by default it will be diskstats from container, not Unraid)

I don't understand what @RBK-Serwer wants to tell. @RBK-Serwer could You please make a thread on Polish Support Forum? https://forums.unraid.net/forum/82-polish-polski/

@Sinister Crayon@professeurx did You create database and database user before You installed Bacula-Server? I've checked on my lab and everything working ok. I've checked postgresql13 postgresql14 and postgresql15 (clean install) That was about curl-101 error with sqlite3 version, but could also happend in postgresql version. Not everything should work in the contenerized version. If you are aware of containerization, you probably know that its purpose is not to run multiple applications within one container and it should not be done. This is necessary here because Unraid out of the box does not support docker-compose. For the same reason, I did not include the Postgres database server in the image like most developers, and for a similar reason it has not been possible to use the MySQL database for a long time. Bacula is a large harvester and should not be run in a container, if it were fully supported Bacula developers would have released images for this purpose long ago. This solution was created for my own needs of backing up data from Unraid virtual machines and several raspberrypi. It was also an alpha project before implementing Bacula Web and later Bacularis in my production environment at work, moreover, it was also my professional development project, for my own satisfaction, and since I put some work into it I felt like sharing it, as you can see, I shared the code on github to which I shared a link, if you want to develop it, please go ahead. For myself and the needs of my Unraid Server, I would run bacula in Docker without UI, because I don't need it, it introduces unnecessary chaos in the configuration files. Of course, you can share ssh, move all configuration files outside the container, only then 90% of people can mess up more and expect help. Just like 90% of users will use 10% of this application's capabilities. What you may want to achieve with this image should be achieved by installing Bacula Server at least in a virtual machine, because then people are surprised that a some process has crashed.

You are doing something that containerization is not intended for. If you need to extend the image with additional functionalities, use the image as a source, install packages, configure and thus use a new image, or use the image creation code and modify it to suit your needs, the image code is published by me on github. Additionally, by modifying the API settings, you may corrupt your copy of the image and require help later (when creating the image, I cannot predict what you will modify, and after updating the image, your settings may be overwritten). Did you provide the ip or hostname by entering DB_Host in the Bacula-Server parameters? Found my mistake, it will be resolved at the next image update. Thanks for the feedback

I don't know what are You talking about. What You put before installing bacula? If You choose postgresql version, You need only to create database user and database in postgresql container/server, If You choose sqlite3 You do not need to do anything exept run bacula container. You must remember that if you tried to install the postgresql version and then want to install the sqlite version, I recommend clearing the paths on the unraid disk where the configuration files are installed. The container installed from scratch updates all passwords - user data, this was tested before releasing a new version of the image. If you want to connect the client to bacula, all you need to do is install the bacula client on the machine that is to be the client, usually called bacula-fd, and then add a new client in the WebUI Bacula-Server. Where do You install client, what are You talking about? What API deploying? What did You tried to do? You do not need to install any API on target machine. You need only bacula-fd. Everything You could configure from Bacule-Server WebUI panel. The exception is the backup of all types of databases, which requires writing a script on the client machine that will dump the database and Bacula will save the dump. The second option is to create remote access to the database and perform the dump from the Bacula server.

Google translate @JonathanM i don't know if there is anybody who can help about bacula image. @professeurx did You update from previous version or this was new installation? Which postgresql server version did You use?

which version did You use? postgresql vs sqlite3 and 11 or 13 maybe 15 ?

You can check if docker is showing unhealthy status and try this kill $(pidof php-fpm7.4) $(pidof apache2) Bacula15 will be available within 1 hour (without s3/amazon plugin yes) There is a problem with s3 plugin and AWS plugin need some additional work. postgresql13 or above is needed for Bacula15 postgresql edition.

@murkus sorry for the delay in replying. There are some known problems with compilation bacula v13.0.4 with S3 driver. Bacula released version 15.x Added Amazon Cloud driver (in replacement of libS3 cloud driver). After the first tests, it doesn't need many changes, but I still want to test the update process to avoid possible problems. Please be patient for a moment