DGMayor 88 Posted February 29, 2016 Share Posted February 29, 2016 I'm not a Linux user, but just wanted to say good find. I'm sure Luke will get it squashed in no time. Link to comment Share on other sites More sharing options...
MSattler 387 Posted February 29, 2016 Share Posted February 29, 2016 I'm not a Linux user, but just wanted to say good find. I'm sure Luke will get it squashed in no time. So I guess it probably is an issue then..... Link to comment Share on other sites More sharing options...
CBers 6766 Posted February 29, 2016 Share Posted February 29, 2016 So this is a Windows and Linux issue then ?? What are the security implications, if any ?? Link to comment Share on other sites More sharing options...
anderbytes 139 Posted February 29, 2016 Author Share Posted February 29, 2016 Not a security issue... but related to performance, availability and stability, instead. As long as I could read the feedbacks, only Linux users are experiencing this.... I don't even know if all Distros are impacted. 1 Link to comment Share on other sites More sharing options...
Redshirt 1487 Posted February 29, 2016 Share Posted February 29, 2016 I'm seeing it in windows, although there could be other factors as well. 1 Link to comment Share on other sites More sharing options...
anderbytes 139 Posted February 29, 2016 Author Share Posted February 29, 2016 I'm seeing it in windows, although there could be other factors as well. Uh oh... in Windows, too?? I guess its worse than I imagined Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted February 29, 2016 Share Posted February 29, 2016 I can confirm I suffered this issue in both Ubuntu 14.04.4 x64 and Debian 8.3 x64 As I mentioned in my other thread the issue is intermittent and for me takes sometime to reproduce. Ie I can go for a week + without any issues. When it does occur I lose all access to the server via the external https url. The local non https url still works fine. Someone mentioned it maybe a firewall issue. I doubt this but for completeness my server is behind a pfsense firewall Link to comment Share on other sites More sharing options...
anderbytes 139 Posted February 29, 2016 Author Share Posted February 29, 2016 I can confirm I suffered this issue in both Ubuntu 14.04.4 x64 and Debian 8.3 x64 As I mentioned in my other thread the issue is intermittent and for me takes sometime to reproduce. Ie I can go for a week + without any issues. When it does occur I lose all access to the server via the external https url. The local non https url still works fine. Someone mentioned it maybe a firewall issue. I doubt this but for completeness my server is behind a pfsense firewall I'm not behind a firewall, and it happens to me. Depending on your use (and if your restart your server often), you may take more or less time to notice the issue. I guess it may be related on how you close your client. Sometimes I close the browser tab, other I sign out, other some crash happens... lots os variables. Same thing to Android : several ways of closing (properly or not) the app and the connection. Luke said he would do something to the next version of dev/beta.... Let's wait Link to comment Share on other sites More sharing options...
Redshirt 1487 Posted February 29, 2016 Share Posted February 29, 2016 I just retested to make sure I wasn't lying when I said I was affected on windows. I went through all my Android devices and Force Stopped any clients. Waited about 10 min and ran netstat with no arguments from the windows CL. Active Connections Proto Local Address Foreign Address State TCP 127.0.0.1:51109 GALACTICA:51110 ESTABLISHED TCP 127.0.0.1:51110 GALACTICA:51109 ESTABLISHED TCP 127.0.0.1:60926 GALACTICA:65001 ESTABLISHED TCP 127.0.0.1:65001 GALACTICA:60926 ESTABLISHED TCP 192.168.0.15:445 CAPRICA:60747 ESTABLISHED TCP 192.168.0.15:8096 192.168.0.18:44582 ESTABLISHED TCP 192.168.0.15:8096 192.168.0.18:59715 ESTABLISHED TCP 192.168.0.15:39901 5.39.92.137:43450 TIME_WAIT TCP 192.168.0.15:39901 5.39.92.137:50297 TIME_WAIT TCP 192.168.0.15:39901 S0106602ad06ef6bb:54826 TIME_WAIT TCP 192.168.0.15:39901 S0106602ad06ef6bb:55122 TIME_WAIT TCP 192.168.0.15:39901 16-235-201-31:42190 TIME_WAIT TCP 192.168.0.15:39901 16-235-201-31:42245 TIME_WAIT TCP 192.168.0.15:39901 16-235-201-31:42445 TIME_WAIT TCP 192.168.0.15:39901 16-235-201-31:42510 TIME_WAIT TCP 192.168.0.15:39901 37.60.74.211:61449 ESTABLISHED TCP 192.168.0.15:39901 37.220.35.196:51989 TIME_WAIT 192.168.0.18 is my Fire TV Stick. Before I force stopped all the clients, 192.168.0.18 had over 3 dozen ports "ESTABLISHED". So the Android clients are holding ports open and requesting new ones endlessly in my situation. Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted February 29, 2016 Share Posted February 29, 2016 Depending on your use (and if your restart your server often), you may take more or less time to notice the issue. I don't restart the server unless I have to. This being the main reason for a restart, otherwise its an update. I mainly use either Kodi plugin or the web browser. I rarely use the android client as it does not work externally - however, I haven't tested this theory, but it seems to be after i attempt to use the android app (ie use it locally where it works) that I then get this issue. I will investigate further this week Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 4, 2016 Author Share Posted March 4, 2016 @@Luke, have you implemented that keep-alive adjust in the new latest stable ?? I've updated my Emby yesterday... and I'm still experiencing the ports-storm... Just to know Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 4, 2016 Share Posted March 4, 2016 The only time i am seeing any issues with this is after I have attempted to use the android app for something. Its fine up to and including the point of using the android app. then 90% of the time its the day after that i run in to this issue. Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 4, 2016 Author Share Posted March 4, 2016 The only time i am seeing any issues with this is after I have attempted to use the android app for something. Its fine up to and including the point of using the android app. then 90% of the time its the day after that i run in to this issue. From what I've tested... 90% of ports are from the Android App... and 10% from other connections. Yesterday I did the following test: - Close all clients (Web, TV, Android, etc...) - Restart Emby Server - check NETSTAT (no connections. OK) - Run Android App, and do some browsing (I used 4G external connection, via https) - check NETSTAT (connections are there, estabilished, OK) - Disable 4G in phone - check NETSTAT (same connections are there, and will be there forever... :-( ) If instead I close Emby App in Android and my 4G is on... the client managed to tell the server to close the ports. If my network drops (or if I turn off 4G), server will never know when to close. In theory... if a TV or desktop is connected and also loses connectivity abruptely... the same thing will happen with them. @, Can you reproduce that? Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 4, 2016 Share Posted March 4, 2016 From what I've tested... 90% of ports are from the Android App... and 10% from other connections. Yesterday I did the following test: - Close all clients (Web, TV, Android, etc...) - Restart Emby Server - check NETSTAT (no connections. OK) - Run Android App, and do some browsing (I used 4G external connection, via https) - check NETSTAT (connections are there, estabilished, OK) - Disable 4G in phone - check NETSTAT (same connections are there, and will be there forever... :-( ) If instead I close Emby App in Android and my 4G is on... the client managed to tell the server to close the ports. If my network drops (or if I turn off 4G), server will never know when to close. In theory... if a TV or desktop is connected and also loses connectivity abruptely... the same thing will happen with them. @, Can you reproduce that? Will double check and confirm this when I get home tonight but I'm pretty sure thats what I was seeing before. It tends to be only me that uses the android App. If I don't use it for a week + I have no issues with being able to get remote access Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 7, 2016 Share Posted March 7, 2016 This issue may be happening tome more frequently than I realised. I have had the server restart plugin installed restarting my server every 24 hours. Will disable and see if the problem is more obvious Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 8, 2016 Author Share Posted March 8, 2016 Any news on that? If devs are watching this thread.... in which stable a fix can be expected? I can't install beta because I installed Emby via OMV plugin system. Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 8, 2016 Share Posted March 8, 2016 So far no. I have turned off the server restart plugin but i am yet to experience the issue. The only thing I haven't done since the last restart was try to use the android app for anything. I will watch something via the emby android app tonight and see how long it takes to become inaccessible Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 8, 2016 Author Share Posted March 8, 2016 OK. Just hope the devs are taking this seriously... even if Linux users aren't the mainstream users for Emby. Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 9, 2016 Share Posted March 9, 2016 OK. Just hope the devs are taking this seriously... even if Linux users aren't the mainstream users for Emby. Agreed Have any of the devs been able to reproduce this? Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 9, 2016 Author Share Posted March 9, 2016 (edited) Agreed Have any of the devs been able to reproduce this? First thing is to summon them (@@ebr and @Luke) to the topic... I wish at least my posts #36 and #38 could be answered. And the most important of all: even if one of them (@Redshirt) could reproduce the problem IN WINDOWS version of Emby (post #34). Can't tell for sure if it is a specific ANDROID client-side issue. If it is the case, move the topic and please try to make a fix. Edited March 9, 2016 by anderbytes Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 16, 2016 Author Share Posted March 16, 2016 No news at all?? Hellooo? [echoes in the room] Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 17, 2016 Share Posted March 17, 2016 Yep still having the issue. Again it only manifests after i use the android app. To confirm this I will not use an android app from any device for the next 2 weeks and will see if the issue appears Link to comment Share on other sites More sharing options...
runtimesandbox 152 Posted March 24, 2016 Share Posted March 24, 2016 Well a quick update, I haven't used the android app for the last week and I so far have not experienced this issue Link to comment Share on other sites More sharing options...
anderbytes 139 Posted March 24, 2016 Author Share Posted March 24, 2016 Certainly there is an existing issue related to Android App and server sockets. Each time Android app connects, there's a 50% chance that this connection gets broken somehow, leaving zombie sockets up, that use resources and won't be automatically liberated. The easiest way to see this is to look at the number of open ports in server, comparing to how many devices are really connected. Sometimes even with few ports open, somehow Android manages to make server "stutter" That's really sad :-((( After next server stable, I'll be putting all my efforts in reporting Android sh** if they are still there. Link to comment Share on other sites More sharing options...
MSattler 387 Posted March 24, 2016 Share Posted March 24, 2016 Certainly there is an existing issue related to Android App and server sockets. Each time Android app connects, there's a 50% chance that this connection gets broken somehow, leaving zombie sockets up, that use resources and won't be automatically liberated. The easiest way to see this is to look at the number of open ports in server, comparing to how many devices are really connected. Sometimes even with few ports open, somehow Android manages to make server "stutter" That's really sad :-((( After next server stable, I'll be putting all my efforts in reporting Android sh** if they are still there. Dust in the wind..... all we are is dust in the wind........ Inevitably this is what happens when a free product becomes a paid product. There are not enough paid developers to keep up with demand, but to hire more developers mean the paid developers take in less cash. This in turn then means that volunteer's are less likely to help for free, when they see the other guys making a ton of money off of it. I see this as the future of Emby. Power and greed corrupts, software development is no different than politics =) Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now