BCinBC

Members
  • Posts

    32
  • Joined

  • Last visited

Posts posted by BCinBC

  1. Is there any documentation on how to configure the Watchdog timer? I have searched the forums and looked briefly through the system options. I'm not sure I want to enable in the command line if it's supported somewhere else.

     

    Quote

    Linux kernel:

    version 5.10.21

    CONFIG_WATCHDOG: Watchdog Timer Support

     

  2. 10 minutes ago, wgstarks said:

    There is already a bug report-

    https://forums.unraid.net/bug-reports/stable-releases/dockers-wanting-to-update-but-dont-in-the-end-r618/

     

    The fix will be included in 6.8, but in the meantime you can make sure that CA Auto Update is current. It will apply the fix automatically.

    Thanks for that! Now I know why it got "fixed!"

     

    I have had CA Auto Update running for ages, but once I did the reboot maybe the the fix kicked in?  I've been up about 6 days now and the issue is certainly gone now. 

  3. 23 minutes ago, queen-adreena said:

    Is it normal to be getting updates for this docker every day... sometimes a couple of times a day?

    I found, that recently, all kinds of dockers were getting Zero Byte downloads and non-removable updates. i.e. You'd update it, it wouldn't really do anything but restart the container. Then when you checked it again, it wanted another upgrade.

     

    I rebooted the server and this went away. If it happens again I'll definitely file a bug report. It had been up for maybe 45+ days?

  4. I tried to install this on my server so that I could remotely access various dockers. However when I activated ZT the main IP address of my management interface disappeared. The only way I could access the server's management interface and all of the various devices on it was only by using ZeroTier even within my own network. 

     

    Is it supposed to do this? I would like to be able to connect to the server within my own network as well as be able to remotely access it with say an Android Phone.

     

  5. 7 hours ago, grither said:

    i had similar problems.  recent plex versions have had problems.  

     

    i would remove docker, and reinstall from template but using version image.png.6039c0805789e3ec564918e173b5d443.png in the key 3 field.  appears to be the last stable version.  my two servers are running this version without problems

     

     

    I found that changing the docker networking mode from Host to Bridge and manually mapping all of the ports needed, It's been rock solid ever since. As soon as I advanced to the new version that had the new network code it was unstable. This seems to affect only a very small group of people but I see it hit the forum every few weeks.

    From My plex forum post: https://forums.plex.tv/t/plex-server-not-worked-since-1-13-4-5271-200287a06/312196/12?u=bobertc

    Quote

     

    I’ve been held back to 1.13.4.5271 for what feels like months. I tried something yesterday I saw on Reddit.

     

    On my unRAID (docker host) box I changed networking from host to bridge. I then added the following port mappings manually:

    Changing the TCP/UDP settings to match the Docker recommended ports from:https://github.com/plexinc/pms-docker:

     

    1900, 32410, 32412, 32413, 32414: UDP
    3005, 8324, 32400, 32469: TCP

     

    Then updated to the latest version of Plex. Available inside and outside, no pauses or disappearing server!

     

    So far so good

     

     

     

     

     

  6. On 9/20/2018 at 10:48 PM, bonienl said:

    Do you mean "br0" or did you define your own custom network "bridge.0" ?

    The newer version of Docker included in Unraid 6.6 has changes in the networking part. We adopted so networks generated by Unraid continue to work as before, but there is no guarantee for custom networks made by the user.

    Custom: br0
    Needed for Unifi, & PiHole.

     

    It also seemed to kill Splunk, which is just a Bridge connection.

     

    So how does one create a new External IP for a docker with 6.6?

  7. Did anyone else have their bridge configuration disappear? I had a few dockers with fixed IP addresses and they suddenly lost Bridge:0. I'm assuming that this happened with the new bridge utils I saw mentioned in the package updates.

     

    I had to undo and redo the settings after backing down to 6.5.3. 

     

    Love the new UI look though.

     

    I posted a bug report.

  8. 9 minutes ago, CHBMB said:

    Not if it's a known Plex issue, I mean I guess it could be a Linux Plex issue, but it's nothing to do with the fact it's running in a docker container.   Why not have a look at the Plex forums and see what's being said like @aptalca suggested.  Might highlight the issue with more clarity than we can.

    It also happened to me on an Ubuntu install on an ESXi server, completely outside of docker. I've been running Plex in my environment for 4-5 years without issue on all kinds of platforms, mostly Linux. This version, I cannot go to as the server cannot phone home to Plex properly. I'll try it on Bare metal if someone really would like me to. It may be "environmental" too. I was thinking of trying it on a second network under pfSense instead of the Ubiquiti ER-X it is now.

  9. 6 hours ago, subagon said:

    Other than posting to the Plex post forums, is there a better method to contact Plex? Is there a way to open a bug report, etc?

     

    It was on this thread that someone said Plex had been informed of this problem. Searching the Plex forums I found one related thread (maybe more but hard to find), but it doesn't contain any insight other than downgrading your Plex server. I'd be happy to add my voice to whatever was send to Plex about this issue if someone could point me to that discussion.

    https://forums.plex.tv/t/plex-server-only-online-for-30-seconds-every-7-minutes-or-so/293805

     

    Sorry link isn't working on my phone client. 

     

    That's what I opened with Plex. 

     

     

  10. 5 hours ago, saarg said:

     

    We are not allowed to post the version numbers here as it might give you access to something you didn't pay for. I haven't restarted my Plex container in some time, so everything is working here. Or maybe not, as I'm not using Plex often.

    The container contains the latest version at the time of build, but if you restart our container, it will pull the latest update if there are a newer one. For that to work, you need to either set it to latest or public in the VERSION variable.

    If you want a specific version, then specify it in the VERSION variable. And yes, it works. Just tested. You can use the tags on the plexinc dockerhub page in the VERSION variable.

     

    I do understand about the version, I'm not trying to bypass anyone's eligibility (I have a lifetime PlexPass) I'm just passing along what I was recommended by a Plex support person, they said to fix this issue, revert to 1.13.2

     

    https://forums.plex.tv/t/plex-server-only-online-for-30-seconds-every-7-minutes-or-so/293805/35?u=bobertc

     

    ChuckPATeam Member
     

    Regarding the “Lost Internet Connectivity” issue we’re all seeing.

    1.13.5 introduced the ability to bind to an adapter. That works great.
    Unforseen side effect is it disturbed MyPlex.
    Engineering is working on that now and , AFAIK, will be in next update. BUT, i’m deep in NAS work so don’t have the absolute latest info on it.

     

     

  11. 47 minutes ago, saarg said:

     

    That will not work. We do not tag the container versions as the plex version. What you have done, will just make an error when it tries to pull the container.

    @saarg

     

    You are correct, mea culpa. I did this on Plex's docker image instead. I needed a fix after being down after 3 weeks, I've tried for weeks to find a fix. Plex has messed up the My Plex code and is causing a lot of grief for specific individuals. Reverting to 1.13.2 is the recommended fix at the moment for those that have Plex stability issues.

     

    I actually did this on a different docker Image. In my haste of the excitement of finding a fix, I didn't realize that I had switched images to:

     

    • "plexinc/pms-docker:1.13.2.5154-fd05be322"

     

    That pulled correctly for me, but I would definitely have preferred to use  linuxserver.io's version instead.I tried setting the VERSION variable to a previous version but I was unsuccessful to find a working one.

     

    If you followed my stupid advice, you'll want to go back into your "Add docker" remove my advice and put in the "repository" field back to just "linuxserver/plex" that should pull the correct image again, but not fix your version number.

     

    @saarg If you actually have the Linuxserver.io's 1.13.2.5154 image number, I'd gladly try it in the VERSION field.

     

     

  12. 14 hours ago, subagon said:

    I see that there's a update to the Plex docker. Anyone know if it fixed the connection timeout issue?

     

    Don't do this, this was done on a different Plex image.

     

    If you followed my stupid advice, you'll want to go back into your "Add docker" remove my advice and put in the "repository" field as just "linuxserver/plex" that should pull the correct image again, but not fix your version number, which is why you're looking for a fix in the first place.

     

    Add ":1.13.2.5154-fd05be322" After 'linuxserver/plex' 

    So it looks like this 'linuxserver/plex:1.13.2.5154-fd05be322'  then apply / done. 

     

    I then removed the container and removed all of the files, tick the checkbox when you remove it (This does not remove the data. Remember that docker stores your user data in the appdata area). This is just to make sure there were no residuals. Then  re-install the docker from your own templates (it should now have saved the value you added above) 

     

    It's been running for about 8 hours with no connectivity issues. 

     

    Also make sure that you are Not automatically updating your docker. 

     

    I noticed that once I forced it to this version it no longer says it has an update. 

     

  13. 10 hours ago, aptalca said:

    With reverse proxy, you tell the plex server what the public domain is, and you can turn off the built in remote access service. Clients connect through your domain and letsencrypt/nginx container handles the ssl and proxy

    6

     

    @aptalca will this eliminate any network throttling and have the system assume these are all local connections?

  14. 3 hours ago, aptalca said:

    For all those having issues with plex remote access, did you try reverse proxying? I had intermittent connection issues until I reverse proxied. No more connection problems. 

     

    https://github.com/linuxserver/docker-letsencrypt/blob/master/root/defaults/proxy-confs/plex.subdomain.conf.sample

    @aptalca The issue is with internal and external connectivity. But this is a great piece of info!

  15. 2 hours ago, wgstarks said:

    If you want to revert to a previous version you’ll need the version number.

    https://lime-technology.com/forums/topic/40463-support-linuxserverio-plex-media-server/?do=findComment&comment=607587

     

    That's a neat trick.  Look in this directory for previous reports:

     

    /mnt/user/appdata/PlexMediaServer/Library/Application Support/Plex Media Server/Crash Reports

     

    However, mine's pretty much empty because I've re-installed plex about 13 times with 3 different distros and also blew away the docker file and app file.

     

     

  16. 7 minutes ago, trurl said:

    The Docker and Github links in the first post explain how to install a specific version of Plex.

    I know that you can specify a build name that the dev has done usually like plexpass, beta, developer etc. But I didn't think it was that trivial to specify a Plex (or other) version number. Thanks, I'll check it out when I get home. Answers are brief as I'm on the train home. I Need 1.13.2, apparently. 

  17. 19 minutes ago, subagon said:

    I've been having this same issue. Started about 10 days ago. No changes other than updating to latest docker. My logs file look just like this one, except for IP's. Did you resolve the issue? My Plex server is unreachable and the family is getting violent.

     

    Edit; It's back online right now, but it goes up and down. Same pattern for last 10 days, but seems to be getting worse.

    If it comes up for a few seconds every few minutes. Plex is aware there is a problem, not sure when it will be fixed. See my previous post on this issue. I have a link to the Plex forum with my post on it. 

     

    The fix, for now? Go back to the previous version. (and no, I don't know how to do that. I wish I could too) 

     

     

  18. Just now, cammelspit said:

    The difference here is this virbr0 I made myself from the command line as part of the tutorial. I should be able to remove this specific network, shouldn't I? It is not part of the docker stuff by default.

     

    Oh, so if you want to change networking and adding bridges, if I recall, you need to spin down your array before you can change the advanced network settings to allow you to add virtual NICs and bridges. 

     

    I spent so much time over the past 10 days with this, I've forgotten what I did. I've just turned off Plex until they come out with a new relese. I've thrown in the towel and am using unRAID with my KODI boxes. :)

  19. 5 minutes ago, cammelspit said:

     

    I think you are on to something. First off, I originally set up this docker when the server was on the 10.10.10.0/24 subnet, same as my current machine so that's prolly why it worked. And with what @ijuarez and @Hoopster said, it seems that it is binding to some subnet with 192.168. subnet. Ok, so I was following Spaceinvader Ones's tut on nextcloud. In his tutorial, he has you make a new network, low and behold it is set to a 192.168. subnet! So this tells me plex is binding to Vibr0. I have since moved the server back to the 10.10.10.99/24 subnet so that shouldn't be an issue right now. However, in unraid, I can't use the GUI to remove virbr0 route. vibr.thumb.PNG.f855c84086ced4aa89af7734509ade3e.PNG

     

    My issue now is I simply don't know the commands to remove it entirely. I decided after the fact that I didn't want to have any of my docker going through virbr0 since it made them inaccessible locally due to the way my network was set up. I COMPLTELY forgot about this too as I am not using it and sort of haven't thought about it in a bit.

     

    I would be willing to put money on this being the problem. Does someone know the commends to remove this virtual network?

     

    Docker natively has a bunch of networking, subnets, etc. that you're not going to be able to remove. Plex tried to make the code that would allow you to specify which network interface to use. While you *CAN* specify that, the code underlying this detection method broke in the last release. They're working on it, and so I wait, I can't get any Plex server to run in my environment under Docker or Linux, ATM.

     

    A quick test is to launch a web client and try to connect to the server. Mine literally is available for 30 seconds every 5 minutes or so. That's the acid test I found. It won't show up initially for a few minutes too.

     

     

     

     

  20. I'm going to add this here, according to @ChuckPA at Plex.tv 

     

    https://forums.plex.tv/t/plex-server-only-online-for-30-seconds-every-7-minutes-or-so/293805/30?u=bobertc

     

    Quote

     

    Regarding the “Lost Internet Connectivity” issue we’re all seeing.

    1.13.5 introduced the ability to bind to an adapter. That works great.
    Unforseen side effect is it disturbed MyPlex.
    Engineering is working on that now and , AFAIK, will be in next update. BUT, i’m deep in NAS work so don’t have the absolute latest info on it.

     

    5

     

    This completely messes up the docker experience, I have a sneaking suspicion we're going to see a lot more complaints about Plex and docker implementations until this gets fixed.

     

    I hope I saved someone some time, I've spent way too much time in the past week on this.

  21. I've been bashing my head against the wall for the past week as I too have had my Plex docker container go offline, it stay up for about 30 seconds every 5-7 minutes. I've got a discussion thread on the Plex forums about it. At first I thought it was a network DNS rebind issue (wasted 4 days of my long weekend on that). 

     

    Plex recently has changed how the networking is done, there's also a new Advanced server feature that lets you select which network interface to use that apparently just got added.

     

    Since last week, no container I've used, be it, linuxserver.io, Binhex's plex or his plexpass. All three have the same issue. 

     

    I even fired up my old previous unRAID box, Chimera. (I saw this server name earlier in this thread. :) The current one is named Bellerophon.) I also spun up a full build of Ubuntu 16.04 LTS on my ESXi box and it too (which made me think it was my network) had this problem. I haven't revisited it yet and tried the new watch network interface, but the VM only had one interface which was directly bound to my local network.

     

    Reading the logs as soon as they say "Aug 07, 2018 10:05:02.800 [0x14d5c5bff700] DEBUG - MyPlex: We appear to have regained Internet connectivity." 

     

    The server vanishes and won't appear for another 30 seconds in a few more minutes.

     

    https://forums.plex.tv/t/plex-server-only-online-for-30-seconds-every-7-minutes-or-so/293805/7

     

    Just to be clear, this is on my internal network, the issues of trying to watch it remotely are at the mercy of the server being up or not.

     

    I've tried two of the four interfaces that appear, I've tried it in both host and bridged mode too assigning it an external new IP address, same results.

     

    So before I try firing up a windows box to use plex with (shudder) please, anyone have any ideas?

     

  22. I would suggest using a UPS, the system will shut down cleanly in the event of a power outage. You need to get a signal to unRAID so that it knows that the power has gone out. A Serial, USB or even an SNMP connection via the network can be leveraged. Make sure that the switch you would send this signal through is also on the UPS if you go with SNMP! :) 

     

    I also found that using the Tips & Tweaks plug-in and selecting "ssh,bash" as the "Processes to kill before Array is Stopped:" helped immensely for shutdowns too.

     

    Also, I can confirm, 'poweroff' works nicely for a clean shutdown, just tried it.