danioj

Make sure your bmc is using the latest firmware. Usually different than bios and can be updated through the board's ipmi web interface. I assume you're using the latest version of the plugin. The ipmi.plg from the original post and not the previous ipmitool-plugin. I would delete the /boot/config/plugins/ipmi directory too. So no sensors are selected for display in the footer and no monitoring enabled. No errors in the system log? Are you using local or network based? If local, try enabling network with the ip of your bmc, username and password. If that works there might be a problem communicating with ipmi through the kernel driver. The plugin checks for modprobe ipmi_si success but if that works and there are communication errors, freeipmi doesn't have an adjustable timeout for local connections like ipmitool did. I think it's 20 sec. Try running the commands from the command prompt (ipmi-sensors, ipmi-sel, ipmi-fru) and see if they lag. Also if you install the plugin and it's lagging run "ps aux | grep ipmi" and "pd aux | grep php" from the command line and see if anything from the plugin is holding up the webgui/emttp. That's all I can think of right now. Let me know. Edit: And I hope everything's ok thank you for the kind words, it is now! In and out of hospital for the past month with various infections. Think twice before getting your wisdom teeth out is all I can say! Anyway .... I had missed the fact there was a new plugin. I was still using ipmi-tool. Easy fix, following your instructions. Removed the plugin, removed the dir from the flash, installed new plugin from CA. Reconfigured. Works perfectly. As an aside, nice upgrade efforts. It looks very professional. Excellent work!

I am sorry to say that i have found an issue with this plugin. Unfortunately, a few weeks before i went into hospital i found that my webgui was "lagging". What I mean is that the structure of the page would load (i.e. tables and background images etc) but the data in the pages (i.e. drive data, smart data, docker images, icons etc) would not "appear" or the page finish loading for about 5-8 seconds. It wasn't happening on my Backup server but that was not a surprise as the Backup Server has very little in terms of additional features enabled. I have narrowed it down to this plugin. Uninstalled everything loads zipidy quick. Installed, the load of the page lags. I have installed - reconfigured - and uninstalled multiple times - still the same result. I am happy to help you debug this. I would like to install this again. See my Main Server setup in my Sig.

Just edit the go file and add the following to it. chown nobody:nobody /tmp/emby Then when the system boots up it will perform that action on the folder eliminating the need to go in each time and reboot. In addition, it is worth noting that you should keep an eye on what you put into the go file as when you upgrade to future versions it can have impacting effects. I always restore my default go file before upgrading, upgrade then re-apply. folder doesn't exist until the docker restarts and without wait commands in go prior to issuing the chown, the command will fail. Hence why my linked post has a mkdir -p /tmp/emby in it first Ah, now I get it. Couldn't he just create it from the go file too and set the permissions? yeah.... This can be corrected in the Go file so it "appears" to be persistent on reboot. Sorry for being obtuse, I still consider myself an "end user" making my way towards "power-user." Can you, or someone in the community, point me to either the Go file, or to the link of documentation which outlines the Go file (so I can RTFM)? I'm assuming the go file is something other then rc.d scripts? /config/go on your flash drive (/boot/config/go) Standard bash script. I guess you would wind up adding these commands to it mkdir -p /tmp/emby chown nobody:nobody /tmp/emby Oh, I am sorry. RTFP Daniel! My only excuse is I am self medicating pain still! If I could I would Giggle!

Just edit the go file and add the following to it. chown nobody:nobody /tmp/emby Then when the system boots up it will perform that action on the folder eliminating the need to go in each time and reboot. In addition, it is worth noting that you should keep an eye on what you put into the go file as when you upgrade to future versions it can have impacting effects. I always restore my default go file before upgrading, upgrade then re-apply. folder doesn't exist until the docker restarts and without wait commands in go prior to issuing the chown, the command will fail. Hence why my linked post has a mkdir -p /tmp/emby in it first Ah, now I get it. Couldn't he just create it from the go file too and set the permissions?

I only have default settings and it shows for me: System Performance (Pystone)194931 Intel(R) Xeon(R) CPU E3-1… Download folder speed251.5 MB/s (/incomplete-downloads) Complete folder speed199.7 MB/s (/complete-downloads) Used cache38 MB (52 articles) [/quote]

Just edit the go file and add the following to it. chown nobody:nobody /tmp/emby Then when the system boots up it will perform that action on the folder eliminating the need to go in each time and reboot. In addition, it is worth noting that you should keep an eye on what you put into the go file as when you upgrade to future versions it can have impacting effects. I always restore my default go file before upgrading, upgrade then re-apply.

Ok, good to know. Will try it when I'm back at home tonight. When interim question though, is helping me understand users and how authentication works. When I created another second user account (let's call it "bob"), it didn't prompt me for a password anywhere. When i download the profile, and send to my device, how does it know who I want to auth as? While you are making "bob" or have made it, in User Permissions on OPEN-AS webui, to the right you should see a show link, click on that, from that dialog box you can change/set user passwords for VPN client. If the server is running you'll have to go to the top of UI an push the changes to the server. So now in your browser if you go to: https://foo_myIP.orDOMAIN:943/ you put in "bob" and password. This will get you DL's to OpenVPN client (if you don't have it already) and the client config files. Grab your config file and VPN client and go. If you already have your VPN client installed and configured, you can just open that put in bob and password. Unless it's the autorun profile which case your VPN client should autoconnect without username and password. Sorry, I do not believe you are correct. In its default configuration (if at all) there are no options for changing passwords at all in the web-gui of openvpn-as. All user username / passwords are managed by the containers "system" passwords. This is why you have to access the container via the command line to create a user and a password (and also change the admin acc password). Once done, you can enable auto login and then you can log onto the access server gui and download the certificate driven profile. Easy Peasy!

It's on the bottom of the first post.

Good pointer. Done. Much better.

Well, this is the first day I have actually had an hour without any pain. Anyway, as a result I took the opportunity to test the updates I made to the script while I was in hospital living off ice-cream! This is a VERY significant update. Having all that time made me realise just how much was wrong with the damn thing before. I have made many changes. I had to remove the code from the forum now as there is too much for a post, but the file is of course attached. Have a look at the change log. Any comments are welcome. After testing myself I feel it is getting to a point where it might be getting good enough for a plugin. As I mentioned earlier, I am pursuing that now!

Thanks for the kind words. I'm recovering. Made some in-roads into this while bed ridden though!!

Thanks for trying it out. It's good to captures as many individual setups as possible. Agreed. Agreed. I will have to check this again because for me, when the VM's were shutdown it worked perfectly. I am already looking at a plugin for another script I wrote. I have been in hospital this week but am working on migrating this into a plugin already. Even if others don't want it, I want it for myself

Short answer? It depends. May work, may not work. In general, it's not worth the risk of a backup not working. Why do you want to back it up while it's running? To add to what jonathanm said: Copying the image files is ALWAYS ok if the guest virtual machines are stopped. But if they are running its possible that as the copy is taking place changes are made within the guest, leading to an inconsistent copy.

I have no idea what your issue is BUT I have just updated both my servers' plugins and I did not have the same issue.

Thanks. Let me know what you think. I tested it over and over again when I had telnet access in the hospital and if the inputs are similar to as I have put in the example it seems to work just fine. If there is enough interest in it I shall tighten it up.

Thanks for the guide/help danioj, all working perfect now! As for Host mode, not sure why it doesn't work, only other Docker/Plugin I have is MariaDB. It doesn't matter as it is now all working fine in Bridge mode No worries, glad you're up and running.

I think that might be my problem, I can't access the Web GUI at all in Host mode, so had to set to Bridge mode to access. What do I need to do/change to setup in Bridge mode? I'm want to use UDP 1194. Note: I setup a quick Debian Wheezy VM and installed OpenVPN AS, that works fine, but would prefer to use Docker for this service. That could be because you have some port conflicts but without more information I don't know. Anyway, if the goal is to get you running quickly the fact that your setup is like you explain, well, thats good, that means your setup is just like mine is on my Main Server. Setup like this and I promise it works great: Admin and Connect URL (assuming your unRAID Server is called Tower OR has an IP of 192.168.1.2): http://Tower:943 OR http://192.168.1.2:943 External Open VPN Port (assuming your domain name is www.example.com OR Dynamic DNS is my_sample_name.dyndns.org OR IP Address is 111.222.333.444 ): example.com:1194 [uDP] OR my_sample_name.dyndns.org:1194 [uDP] OR 111.222.333.444:1194 [uDP] And don't forget your router config:

Thanks for the explanation. Should I be able to test OpenVPN internally from a machine on the network? (I have the AS Host Name set to UNRAID local IP) I can't seem to get OpenVPN GUI on Windows to connect to the Server: This Docker is set to Bridge with Privileged On, see attached screenshot for a few settings/error: I think you should be able to do this. Note that the Docker is set to run in Host mode. Please see this from the docker hub page linked to in the first post of this thread: https://hub.docker.com/r/linuxserver/openvpn-as/ Now this is NOT strictly true. You CAN operate in Bridge mode AFTER you have setup the application in host mode OR if you have decided which ports the services are going to operate on. What I did was set OpenVPN-AS to run the VPN Server on port 1194 and the Admin / Connect services to run on 943 and tested it worked. Then I switched from Host to Bridge and mapped Container ports 1194 and 943 to the Host. Obviously IF I ever want to change these ports I won't be able to change the ports in the OpenVPN-AS Admin web gui (like I would be able to if I had this setup in Host mode) I will have to change the ports of the Bridge mode setup. You might ask why I did this and it was because I have a bonded NIC setup and this setup wasn't playing well with this Docker in Host mode. In bridge mode it worked perfect. Anyway, I digress. Setup the docker working in host mode and AFAIK you should be able to do what you are trying.

I have just tried it and it is working fine. Are you still having issues?

WARNING - PLEASE CONSIDER THIS A WORK IN PROGRESS. I HAVE TESTED IT ON MY SERVER AND THERE WAS NO ISSUE BUT THAT DOESNT MEAN IN ANY WAY IT IS FREE FROM BUGS/ISSUS SO PLEASE USE AT YOUR OWN RISK UNTIL IT HAS BEEN TESTED FURTHER - WARNING Well, this is what happens when I am in hospital with nothing to do but wait for tests. For those who don't like what I have done, then I am sorry! Anyway, I was discussing the need for backup of Virtual Machine vdisks. I certainly need a method that is not manual and I thought I would write a "Proof of Concept" script to do this for us to see if it might be something that would be needed by others too. Like I said above, this is a proof of concept initially and it is missing most of the refinements that go into good code and a released product. For that reason I would suggest if you don't know what you're doing then don't use this. If there are people who want to test this in their test environment and or give feedback please do the following: - download script from github (link at bottom of the post). - copy the script to your flash drive. - read the notes in the file. - fill in the mandatory inputs (e.g. virtual machine list and backup location). - change optional switches if you wish to. - from the cli execute it like below (for now). sh backup_unraid_vms.sh - write a post and give feedback # this script has been created to automate the backup of unRAID vm's vdisks(s) to a specified location. # this script does not yet run using variables passed from the cli as yet but there is intention to do so if there is interest. # for now the variables below are what are needed to be ammended to have the script run. # Change Log # v0.1 20160330. # initial release. # # notes: initial realease is more concept and to see if there is a need for it. it lacks basic error handling and recovery. # which could be developed later. # v0.2 20160330. # bug fixes. # # changed how the bash script is recognising the failed shutdown flag as after testing even a failed shutdown would start a backup. # changed default mins to wait for shutdown to 6. # v0.3 20160415. # bug fixes and enhancements. # # applied strickter naming conventions to variables. # cleaned up the code, added comments, removed unecessary 'do' loops. # added input validation and verification. # added addition status messages. applied a bit more consistency. # added ability to deal with vm's with names that have spaces in them. # vms now seperated by "new line" not space. # added rsync copy over standard nix cp command. # added backup of vm xml configuration. # added option to add timestamps to backup files. # added script name check for version control. # added option to enable / disable the script. # added option to start vm after failure. # added option to "dry-run" copy operation - note all other functionality is enabled. # added ability to ignore vm's which are not installed on the system. # set defaults for all options to 0. # added guidence for options and inputs. # added constraint to only backup .img vdisks so to skip installation media etc. # changed method of obtaining vdisks list. # fixed issue which had script fail if no vdisks were present. # v0.4 20160416. # administration. # # script name changed to facilitate code being added to github. # bug tracker (order by severity) # <reference> <description> <link to post> <severity> <date added> <by whome> <accepted/refected> <comment> # to do list # core # error capturing and handling -- started (v0.3). # apply stricter naming conventions -- done (v0.3). # code clean up -- ongoing (v0.3) # input validation and verification -- done (v0.3). # logging. # clean up status messages -- done (v0.3). # deal with vm's named with spaces -- done (v0.3). # possible improvements # use rsync to copy the vdisks instead of copy -- done (v0.3). # possible future features # backup vm xml file -- done (v0.3). # have the script run from variables passed in from the cli. # add timestamps to files -- done (v0.3). # add iterations of backups and number of backups to maintain at any given time. # plugin -- started (v0.3). # plugin to-do's # basic structure and requirements -- done (v0.3) # git-hub account # ui validation via menus # .... much much more .... script now available via github: https://github.com/danioj/unraid-autovmbackup.git

Actually, you can just go to the tools page of this plugin and press the Clear button. This will delete any attributes that were added before the files were excluded, and once the attributes are gone the files will no longer be processed. No matter, deleted them now! Thanks though!

A very very BIG +1

Right, that makes more sense. I shall get rid of ALL .DS_Store files and see if I have any issues going forward. Understanding now what the export function is and how external files are used I don't think I need it, I have a full daily backup to restore from which is also being checked regularly AND i doubt I would ever get rot on two files at the same time so even if there was rot I will always have one copy of the file to restore from so I shall not worry about it. Thank you for taking the time to explain that to me

I have come across some behaviour with this plugin that i'd like to see if others are experiencing the same thing as me. I started with a full scan and export, checked to monitor external files and also have an exclusion on .nfo's and apple metadata. My settings are as indicated in the screenshot below: So, I am expecting that once I have scanned and exported that my interaction with this plugin will be limited to any issues it finds on a monthly basis. Is this a realistic expectation? When the scan executed a couple of days ago (27th) I received the following email: Event: unRAID file corruption Subject: Notice [MAIN] - bunker verify command Description: Found 5 files with BLAKE2 hash key mismatch Importance: warning BLAKE2 hash key mismatch (updated), /mnt/disk4/nas/Documents/.DS_Store was modified BLAKE2 hash key mismatch (updated), /mnt/disk4/nas/Documents/Daniel/.DS_Store was modified BLAKE2 hash key mismatch (updated), /mnt/disk4/nas/Documents/Daniel/development/scripts/.DS_Store was modified BLAKE2 hash key mismatch (updated), /mnt/disk4/nas/Documents/Daniel/development/.DS_Store was modified BLAKE2 hash key mismatch (updated), /mnt/disk4/nas/.DS_Store was modified I was expecting that .DS_Store files would be ignored as I have selected ignore apple metadata? Is there something I am missing? The ignore of .nfo files seems to be working just fine. In addition I keep getting emails saying the following: Event: Dynamix file integrity daily update Subject: Notice [MAIN] - Disk 2, Disk 4, Disk 6 needs Export updating Description: 29 non-existing files in export file Importance: normal Why if I have things set up to monitor new files should I have to keep going in and doing a manual export? I am expecting more from the plugin than it is designed to do? Last but not least, when I DO go in there and go to do the manual export I find (each time) that there is not only an export that is required but there needs to be a scan done on one or more disks too. See screenshot below: Now, I can do a Build and Export again BUT feel like I am having to manage this Plugin and keep monitoring the status to ensure that everything remains Built and Exported. Is there anything I need to be doing differently or do my expectations of the plugin need to change? For now Build and Export status is back up to date: EDIT: And now almost 3 hours later, the Build is showing as not up to date for Disk 4. There have been some files copied to Disk 4 BUT I thought that new files were being monitored and things updated accordingly