So on my Plex, the remote access keeps randomly disconnecting. In order to start it again I have to disable it and re-enable it. It will work for a few hours then drop again.
Nothing has changed on my network. This just started within the last week. I verified all the settings are correct. The port is forwarded properly in my router. I have FiOS.
I’ve gone through the help page on their site in regards to NAT and uPNP and it still keeps failing.
I can’t for the life of me figure out why this keeps happening so much. Any ideas?
We can start by looking at your configuration details and also some logs covering period when it was ok and up to time when it is not.
Please be aware that the indicator on the remote access server settings page may not be accurate and sometimes it may give a false indication - eg saying it is unavailable when it actually is. This is due to some timing issues and will be addressed
So ensure your IP address of the server is a DHCP Reservation or static IP outside DHCP range but of course within the subnet range
ensure you only have one network interface active - so if you have wired and wireless, disable one of them and make sure the port forward in the router goes to the right active interface - which should be a fixed IP - eg DHCP Reserved
ensure debug logging is enabled on the server
Settings / Server / General / Show Advanced
When it fails and you are no longer available to connect to it - eg through cellular from mobile Plex app on android or ios or windows phone, ore remote wifi, then capture server logs again
Note down time of failure / screenshots
and attach here together with configuration details
so screenshots of the port forward screen
screenshot showing the DHCP Reservation ipconfig /all
screenshot of server settings remote access with show advanced
confirmation that the router public wan IP Address (router status page) is the same as the IP displayed by http://canyouseeme.org
Well this is nice to see. I am on Linux myself (not windows like OP) but have the same problem. I am disheartened that the same answer was given of Oh it just sometimes looks like that. No, It really disconnects. At least you tried an answer here, where my question in the linux side got an answer from chuck that was basically how the programmers are just so busy they can’t work on this now, and how he felt it was ok since in home was the most important thing. Well it isn’t. It also isn’t our setup.
OP and I both use Fios just different OS, but the same problem. I walk away and come back - boom no remote access, just like OP. I go ride a bus and whoops no access. I have to remote into the network to get a web browser to re-enable. Don’t insult our intelligence by saying it just looks offline…NO, IT GOES OFFLINE. I have done Static IP. Router shows proper port forwarding, but in another couple of hours i’ll be offline again. Ridiculous. Saying it’s a timing issue that you’ll get to is also ridiculous. Just friggin get to it.
I know I pay for plex pass partly for the remote access benefits. Why am I (and others) paying for something that you people don’t seem to have a priority on? You offer something for pay, those features should get a little priority over the free stuff, or why am I paying?
I’m continuing to try things in the router as I’m sure the OP is, but it definitely seems like plex’s fault. And if its not the fault of it, per se, it’s plex’s problem. Using UPNP (i assume) you should be able to keep ports open. Just either way could you guys try and pretend like you care that your paying customers are being super inconveienced like this? I’m half tempted to scrap it and go with one of the open source linux alternatives. Not as pretty or convenient, but I can be damn sure they will be live on the internet.
And with that I end my rant. Sorry it got so long, it’s just very frustrating.
@neshaboi said:
Well this is nice to see. I am on Linux myself (not windows like OP) but have the same problem. I am disheartened that the same answer was given of Oh it just sometimes looks like that. No, It really disconnects. At least you tried an answer here, where my question in the linux side got an answer from chuck that was basically how the programmers are just so busy they can’t work on this now, and how he felt it was ok since in home was the most important thing. Well it isn’t. It also isn’t our setup.
OP and I both use Fios just different OS, but the same problem. I walk away and come back - boom no remote access, just like OP. I go ride a bus and whoops no access. I have to remote into the network to get a web browser to re-enable. Don’t insult our intelligence by saying it just looks offline…NO, IT GOES OFFLINE. I have done Static IP. Router shows proper port forwarding, but in another couple of hours i’ll be offline again. Ridiculous. Saying it’s a timing issue that you’ll get to is also ridiculous. Just friggin get to it.
I know I pay for plex pass partly for the remote access benefits. Why am I (and others) paying for something that you people don’t seem to have a priority on? You offer something for pay, those features should get a little priority over the free stuff, or why am I paying?
I’m continuing to try things in the router as I’m sure the OP is, but it definitely seems like plex’s fault. And if its not the fault of it, per se, it’s plex’s problem. Using UPNP (i assume) you should be able to keep ports open. Just either way could you guys try and pretend like you care that your paying customers are being super inconveienced like this? I’m half tempted to scrap it and go with one of the open source linux alternatives. Not as pretty or convenient, but I can be damn sure they will be live on the internet.
And with that I end my rant. Sorry it got so long, it’s just very frustrating.
Yeah, it definitely is not a problem on our end. Nothing changed since I got to FiOS months ago and all of a sudden this started happening. This is most definitely on their end. I am going to try and downgrade the software and see if the older version is going to work.
Update and reply to MThomas - My own researching seems to show that the UPNP of the plex server may be a little deficient. It looks like it’s not staying on top of keeping the upnp request up to date. I went ahead and tried to make a static port map in the router. Thus far it looks like it might be working. So I suppose we can just not trust the UPNP system of plex and do the port forward. Seems like a big bug (if this fixes it) since it defaults to upnp and doesn’t even suggest something else. I also don’t see an option (i could have missed it) to turn off UPNP so we can do manual.
To the plex representatives on here - First please try and do it on your end, but when someone comes on here with this problem, try suggesting manual port forwarding as a first step, rather than screenshots and logs and other such tripe. This seems like a simple solution or at least first step before getting more technical.
I have a netgear 6400 router and I have issues with UPNP as well and I agree, it’s Plex not advertising the port to the router in a timely fashion (or something like this, I’ve never been that interested in networking). When I try to use port forwarding, I have to choose between “Secured” and “Open” for NAT Filtering and only the “Open” setting allows Plex to port forward correctly (UPNP being off at that point).
I don’t like having to set NAT Filtering to “Open” rather than secured and I’d prefer to use UPNP but it drops consistently every hour or so. I’ve adjusted the advertisement period and the time to live (in hops) but nothing makes UPNP stable, I lose remote connectivity all the time.
Definitely seems like there is a bug with how Plex is implementing UPNP…would love some help getting it working properly here. If it’s the settings with Netgear, please, let me know. My Ceton My Media Center is port forwarding with the NAT set to “Secured” and it works fine…
Ok - I always recommend a port forward over uPnP and to manually specify a port. In this thread I did not mention that as I assumed from this sentence by the OP The port is forwarded properly in my router. that a port forward was already in use.
The timing issues I mention relate to the status indicator and I mention this in case it was just the status that was wrong and remote access was not failing. However, I did also ask for logs to try and see why remote access was lost - if it was failing.
Configurations have to be checked in each case because there have been cases when there were multiple IP Addresses, or double nat or IP address for the server not DHCP reserved
There are also timing issues observed when switching between uPnP and Manual selected port / port forward - but once port forward is in place and to be doubly sure server restarted then it should stay working, If it does not, then it needs to be investigated with logs etc…
Greetings WRXFanatic - I’m always glad to get more confirmation of what my problem is. I (several hours later) can now say that the setup I have with manual forwarding appears to be working. My usual router preference is linksys, but I looked up quick info for yours. Please check out the link herehere and take a look. I hope the link worked, not used to doing them on here.
It looks like the router is pretty straightforward. Don’t do insecure nat or dmz. Follow the directions on that site. First go to plex, settings, server, remote access and you’ll see the internal port (prob 32400) and the external port. In your router for the external port beginning and ending should both be the second port shown in that settings page. Uncheck “use the same port range”. Internal start and finish range should be the same as well (32400 for me), and then your ip.
I hope that’s helpful, as I was running from only the web page and couldn’t see your actual pages.
If you need more help. not sure I can but feel free to ask and hopefully we can . get you up and running as well.
To the PLEX Crew - I think we have pretty convincingly demonstrated that you have a faulty UPNP subsys in your software. You need to look at it, and perhaps increase the advertisement rate or whatever you need to do. What you need to NOT do is send us on wild goose chases with screen grabs and logs, and certainly not tell people that it’s not a priority and you have other things to do, as ChuckPA did to me. If someone tells you the remote access works, then drops and works again on refresh try and believe them and help them solve the problem not deflect back to us. That scenario screams upnp trouble so treat it as such. Very disappointed in plex reps on here, though the software in general is pretty good.
Yeah, it definitely is not a problem on our end. Nothing changed since I got to FiOS months ago and all of a sudden this started happening. This is most definitely on their end. I am going to try and downgrade the software and see if the older version is going to work.
Heya mthomas184 - I am replying here rather than message so others can keep up. I hope the board doesn’t delete this as a double post.
Were you able to solve your problem? Since I Had luck on mine, I’d be happy to try and help you if you would like. Feel free to message me or reply on here so others can read the replies when they have trouble too.
I’m not sure that you can download the firmware on their blasted Quantum routers from fios. As far as I understand on boot and every so often it looks for new firmware and upgrades if you like it or not. I’ve seem people with unrelated problems on the new firmware for the quantum pos, including dyndns and other subsystems. You can’t even set a static ip for a device in your house with this stupid router. But they have you over a barrel with the coax connection for your tv boxes as well. As soon as I get a mocha bridge for myself that router is OUT.
Well I never received a reply to my previous post, so I will try in here. I can confirm that remote access works fine for me on version 0.9.17.3 which is way way back! It has been working without a single disconnect for well over a week. As soon as I update PMS to the latest version… my remote access will only stay enabled for maybe an hour at the most. Then I either have to disable/re-enable or shut down PMS and restart it to get it to come back up. So can someone tell me why it works just fine on this old version (0.9.17.3) but not on any version since?
Ok - I always recommend a port forward over uPnP and to manually specify a port. In this thread I did not mention that as I assumed from this sentence by the OP The port is forwarded properly in my router. that a port forward was already in use.
The timing issues I mention relate to the status indicator and I mention this in case it was just the status that was wrong and remote access was not failing. However, I did also ask for logs to try and see why remote access was lost - if it was failing.
Configurations have to be checked in each case because there have been cases when there were multiple IP Addresses, or double nat or IP address for the server not DHCP reserved
There are also timing issues observed when switching between uPnP and Manual selected port / port forward - but once port forward is in place and to be doubly sure server restarted then it should stay working, If it does not, then it needs to be investigated with logs etc…
Heya Sorry if I was a little harsh, you were more helpful than ChuckPA who just told me how we can all agree internal access is the priority., and the programmers have their own issues to work on. It’s frustrating however., and some of that may have (did) bleed over to you. Yes the OP did say ports were properly forwarded, but you could have asked what that meant. In this router the upnp and manual forwards show up in the same spot. During my troubles I saw that it was forwarded according to the router. Then it would die, and a click on refresh redid it. Unless I looked at the router in those critical minutes, it looked like the ports were forwarded.
I’m just saying that a couple of questions could have saved time and saved everyone from getting screenshots and logs. I’m also saying that this is an issue and needs to be escalated. I refuse to believe that we are the only three people using (read paying for) plex and having this issue. The Fios router sucks, but it can’t be the only one that has issues with how often you advertise the upnp. So now it’s on record that there’s a problem with the plex UPNP compatibility. Or some routers have upnp issues, which still translates to your problem. So you and the company are on notice, so the I didn’t know and we have to troubleshoot from the ground up argument won’t hold water.
Again however, I’m sorry for the angry tone towards you, who are being more helpful than some others. You also answered without matching my tone, so I am double sorry and grateful to at least have your ear. Now we can move on and try to resolve for people
@slvrnblck12 said:
Well I never received a reply to my previous post, so I will try in here. I can confirm that remote access works fine for me on version 0.9.17.3 which is way way back! It has been working without a single disconnect for well over a week. As soon as I update PMS to the latest version… my remote access will only stay enabled for maybe an hour at the most. Then I either have to disable/re-enable or shut down PMS and restart it to get it to come back up. So can someone tell me why it works just fine on this old version (0.9.17.3) but not on any version since?
Heya slvrnblck12 welcome to the discussion.
As you can see on here, it seems that the PLEX people changed their UPNP subsystem somewhere, though gods only know why. It appears that if you can get yourself a static ip, and then do manual port forwarding you can force it to work.
It’s interesting that this thread about the issue is gathering more and more people. There was the OP and then you, me and WRX. Pay attention to that PLEX people. This is an issue that is affecting untold numbers of people, again you’re on notice and can’t claim ignorance any more.
Yeah, it definitely is not a problem on our end. Nothing changed since I got to FiOS months ago and all of a sudden this started happening. This is most definitely on their end. I am going to try and downgrade the software and see if the older version is going to work.
Heya mthomas184 - I am replying here rather than message so others can keep up. I hope the board doesn’t delete this as a double post.
Were you able to solve your problem? Since I Had luck on mine, I’d be happy to try and help you if you would like. Feel free to message me or reply on here so others can read the replies when they have trouble too.
I’m not sure that you can download the firmware on their blasted Quantum routers from fios. As far as I understand on boot and every so often it looks for new firmware and upgrades if you like it or not. I’ve seem people with unrelated problems on the new firmware for the quantum pos, including dyndns and other subsystems. You can’t even set a static ip for a device in your house with this stupid router. But they have you over a barrel with the coax connection for your tv boxes as well. As soon as I get a mocha bridge for myself that router is OUT.
I actually have not yet. I was working all night and am gonna look at it tomorrow and see what I can do. I am gonna check out what everyone is saying and see if I can get it going.
I am disheartened that the same answer was given of Oh it just sometimes looks like that. No, It really disconnects.
So, sometimes it is actually beyond our control. There are thousands of different router models; some are buggy. I currently have a Cisco RV320 and I moved over to manually entering the mapping, because the damn thing would leave incorrect mappings around and I got similarly frustrated trying to figure out why I was losing remote connectivity.
As you can see on here, it seems that the PLEX people changed their UPNP subsystem somewhere, though gods only know why.
Please don’t assume it’s always us responsible for every bad thing I know it might seem like that sometimes, but there are many variables here.
As far as I know, we haven’t changed the UPnP mapping code in quite some time (I messed with it locally for a while when trying to resolve my own Cisco issue, but never managed to get it to manage to make the bloody router work properly).
We double check public IP every hour, and if it changes, we re-publish. So if you are having trouble with the dynamic port mapping (which is always going to be less reliable than a static mapping), I suggest you try a static mapping.
Lots of replies here which is great but hard to follow but I never seem to get answers on other questions related to this so I’m going to continue to try here.
I have a Netgear 6400 Router
I absolutely know HOW to manually port forward and have successfully done it for services like Ceton’s My Media Center HOWEVER, the only way I have got it to work for Plex is if I flip Nat filtering from “Secured” to “Open” on the router settings. To summarize, manual port forwarding all setup properly in router and on Plex server to same specified port, Router set to Nat Filtering “Secured” = Can’t access Plex remotely. If I turn NAT Filtering to “Open” (which is a much less secure firewall setting according to Netgear manual) then I can Plex access remotely via manual port forward. Because I don’t want to set my router to “Open” NAT Filtering, my preference would be to use UPNP and leave NAT filtering at “Secured”.
I have PlexPy setup to send notifications for remote access problems and the entries seem to follow a pattern for the most part all day long… Down 5:41am / UP 6:39am / Down 8:41am / UP 9:39am / Down 12:15pm / UP 12:39pm / Down 7:41pm / UP 8:39pm / Down 9:41pm. I can confirm that if I try to access after getting a notification of being down, I can’t access my server directly from outside my network. Plex logs indicate that UPNP port is refreshed hourly?
Based on the information in #3, I’m left to conclude that the Advertisement Period for UPNP and TTL (in hops) settings on the Netgear Router must be adjusted to keep port settings fresh. I have SCOURED the net and found no good articles or solid information here. Furthermore, my UPNP port mappings table in my router seems to be storing lots of stale records for Plex (currently there are 10 different port mappings in the table for Plex to internal Port 32400).
My current Router advertisement period for UPNP is 30 minutes (router default) and TTL (in hops) is set to 50 (Router default was 4). My research leads me to believe that TTL zero restricts it to the same host, one to the same subnet, 32 to the same site, 64 to the same region and 128 to the same continent; 255 is unrestricted. Please make a setting recommendation for UPNP Advertisement period and TTL (in hops). Lower advertisement times seem to create unnecessary network traffic although I’ve heard of some gamers trying to set advertisement to 1 minute (lowest setting) and TTL to 255 (highest setting/unrestricted). This seems like an overly broad approach and could cause other problems.
So, I’m happy to provide logs or whatever needed to resolve this. I think most UPNP issues are not bugs with the router itself but rather the advertisement period and TTL not working with Plex? If we had more universal settings to recommend, I think UPNP would work well on most routers. Most routers use a default of 30 minutes and TTL of 4 so optimizing Plex settings with this in mind would probably make a world of difference too…
I wanted to do more testing and my results were consistent with my statements above. If I turned UPnP off, manually forwarded a specific external TCP port (set on router and in Plex server settings) to internal 32400, reboot router, close PMS and re-launch, when I would check the remote access in PMS settings it would flip green for one second and then back to red immediately. If I flip Nat Filtering to “Open” it stays green. Now for the crow meal…It’s possible the one thing I hadn’t tried in all this was rebooting the entire windows box after setting the manual port forward whilst leaving NAT Filtering set to “Secured”…So, I rebooted my windows box running PMS…I get a green light now at least and NAT filtering is set to secured. I’ll see how things work tomorrow…So like all things IT related, a reboot may have fixed this…mmmm tasty crow…
I still wish the UPnP settings were better worked out. It’s a slick service and I like the idea of having the external port change frequently.
@WRXFanatic said:
Well, I’ll eat a little crow here…
I wanted to do more testing and my results were consistent with my statements above. If I turned UPnP off, manually forwarded a specific external TCP port (set on router and in Plex server settings) to internal 32400, reboot router, close PMS and re-launch, when I would check the remote access in PMS settings it would flip green for one second and then back to red immediately. If I flip Nat Filtering to “Open” it stays green. Now for the crow meal…It’s possible the one thing I hadn’t tried in all this was rebooting the entire windows box after setting the manual port forward whilst leaving NAT Filtering set to “Secured”…So, I rebooted my windows box running PMS…I get a green light now at least and NAT filtering is set to secured. I’ll see how things work tomorrow…So like all things IT related, a reboot may have fixed this…mmmm tasty crow…
Thanks for the update
@WRXFanatic said:
I still wish the UPnP settings were better worked out. It’s a slick service and I like the idea of having the external port change frequently.
Well that poses an issue for all clients. How soon would they detect the changeover to the new port and how smooth that transition can be. At least with a fixed Port Forward the only potential issue of similar nature then would be any change of the public IP which should be much less frequent
I am disheartened that the same answer was given of Oh it just sometimes looks like that. No, It really disconnects.
So, sometimes it is actually beyond our control. There are thousands of different router models; some are buggy.
(Snip)
Please don’t assume it’s always us responsible for every bad thing I know it might seem like that sometimes, but there are many variables here.
As far as I know, we haven’t changed the UPnP mapping code in quite some time (I messed with it locally for a while when trying to resolve my own Cisco issue, but never managed to get it to manage to make the bloody router work properly).
We double check public IP every hour, and if it changes, we re-publish. So if you are having trouble with the dynamic port mapping (which is always going to be less reliable than a static mapping), I suggest you try a static mapping.
Hey again.
Yes it’s never automatically the software for sure. Some routers are buggy for sure.
Just seems like you are starting from a position that even though it happen to you it couldn’t be the software.
However what we have are several people with issues with different routers… Including you apparently. That points to your software. Deflecting that seems a bit disingenuous. Different people, some different isp’s, different Os’s. Report that it changed after version 0.9blahbah. again that all points to your software. Starting any search elsewhere, with these reports, seems like classic hey look over they tactics. Perhaps were a little harsh but all we get, even from you, is basically how it couldn’t be your stuff, despite points that way. That frankly pisses people off. Hear the frustration in my and others and try to look at what’s happening not trying to throw it back to us.
Again at least part of my anger is at chuckpa blowing me off telling me how he thought we could all agree that internal access was most important and the programmers didn’t have time to deal with us anyway… now others have the and problem and are also mad and not interested in being handled but being heard and resolved
But OK if means so much to you when I get home I’ll undo my manual mapping breaking it for me, wait an hour or so and then send you log when it’s offline again.
Btw please specify the exact log files you want, related to networking. I don’t want to post anything with personal info such as user names or what media I have.
@WRXFanatic said:
Lots of replies here which is great but hard to follow but I never seem to get answers on other questions related to this so I’m going to continue to try here.
I have a Netgear 6400 Router
I absolutely know HOW to manually port forward and have successfully done it for services like Ceton’s My Media Center HOWEVER, the only way I have got it to work for Plex is if I flip Nat filtering from “Secured” to “Open” on the router settings. To summarize, manual port forwarding all setup properly in router and on Plex server to same specified port, Router set to Nat Filtering “Secured” = Can’t access Plex remotely. If I turn NAT Filtering to “Open” (which is a much less secure firewall setting according to Netgear manual) then I can Plex access remotely via manual port forward. Because I don’t want to set my router to “Open” NAT Filtering, my preference would be to use UPNP and leave NAT filtering at “Secured”.
I have PlexPy setup to send notifications for remote access problems and the entries seem to follow a pattern for the most part all day long… Down 5:41am / UP 6:39am / Down 8:41am / UP 9:39am / Down 12:15pm / UP 12:39pm / Down 7:41pm / UP 8:39pm / Down 9:41pm. I can confirm that if I try to access after getting a notification of being down, I can’t access my server directly from outside my network. Plex logs indicate that UPNP port is refreshed hourly?
Based on the information in #3, I’m left to conclude that the Advertisement Period for UPNP and TTL (in hops) settings on the Netgear Router must be adjusted to keep port settings fresh. I have SCOURED the net and found no good articles or solid information here. Furthermore, my UPNP port mappings table in my router seems to be storing lots of stale records for Plex (currently there are 10 different port mappings in the table for Plex to internal Port 32400).
My current Router advertisement period for UPNP is 30 minutes (router default) and TTL (in hops) is set to 50 (Router default was 4). My research leads me to believe that TTL zero restricts it to the same host, one to the same subnet, 32 to the same site, 64 to the same region and 128 to the same continent; 255 is unrestricted. Please make a setting recommendation for UPNP Advertisement period and TTL (in hops). Lower advertisement times seem to create unnecessary network traffic although I’ve heard of some gamers trying to set advertisement to 1 minute (lowest setting) and TTL to 255 (highest setting/unrestricted). This seems like an overly broad approach and could cause other problems.
So, I’m happy to provide logs or whatever needed to resolve this. I think most UPNP issues are not bugs with the router itself but rather the advertisement period and TTL not working with Plex? If we had more universal settings to recommend, I think UPNP would work well on most routers. Most routers use a default of 30 minutes and TTL of 4 so optimizing Plex settings with this in mind would probably make a world of difference too…
Hey not home but a few thoughts. I don’t know what you mean upnp advertisement on the router. As far as I understand the advertisement is how often the program updates upnp. Now that 30 mins may be how often is cleans out “old” upnp entries.
Also I’m not sure your thoughts about TTL is correct. It’s time to live. Measures in hops or seconds depending on your source. Check out Wikipedia time to live. It looks like you were confusing TTL will the network number (/24 to cover a whole class c network such as 192.168.1.0/24.
Really this looks like it could be resolved by Plex advertising upnp more often. Maybe 30 mins. The network traffic from the machine to the router would be trivial.
Sorry if I missed anything ask again if you wish:)
I know it might seem like that sometimes, but there are many variables here.