Media Server :: FMS 3.0 Stops Accepting Connections After 1-3 Days, Restart Required?
Oct 7, 2009
We have a custom application that is running on FMS. Server is dedicated just for fms, there is only one application running on the server, with more instances, average traffic is 3-15 simultaneous users.The problem is,server suddenly stops accepting connections (after 1-3 days from restart), from this point it is impossible to connect to any instance, and fms server must be restarted.When the server crashes, the admin console can see all open instances, can close instances, can restart fms. There is nothing logged, that would explain this sudden failure.Second problem is, cores are connected and disconnected too often, causing client to reconnect (takes 2-5 seconds).
Hardware: 4 x quadcore Intel(R) Xeon(R) CPU E7440 @ 2.40GHz, 32 GB ram
OS: CentOS release 5.3 (Final), kernel: 2.6.18-128.4.1.el5
cpu and mem usage: 1-5% max[code]...........
We have one application in which we are using flash media server 2.0.5. We dont know why it's stop taking connection occasionally at that time we can't enter into the managment console. We get a white screen. When we try to enter user name and password it's does not accpet the credentials. We are not getting anything in the log.Port number 1935 and 1111 of the particular system where FMS has placed are working fine.
I have been running FMIS 3.5 for about 18 months, and recently it started acting up even though I have not made any changes. After a while (sometimes days, sometimes hours), it starts rejecting new connections, but it allows existing connections to keep working. If I refresh my app, then I can't reconnect. The license I have is unlimited. Can anyone shed some light into why this is doing this and what I can do to fix it? Sometimes it happens when only a couple dozen people are online, and today it did it about 4 times with over 1500 people online.
I have a setup where I have some edge servers (as in <Mode>remote</Mode>) that pull content from an origin server. I only want to allow RTMPE connections to the edge server. I'm trying to do this by using the <DisallowedProtocols> element on the origin server. I'm setting: <DisallowedProtocols>rtmp,rtmps,rtmpt</DisallowedProtocols>
in the relevant Vhost.xml on the origin server. However, when I do that, proxying through the edge server stops working. I get a "Connection attempt rejected by FMS server [ Server.Reject ] : Connection failed." from OSMFPlayer (which I'm using for testing) and these lines show up on the origin server's core.00.log file:
I have instaled FMS 2 on local machine and have a video chatroom. I have used:rtmp//localhost/chat as the string and i can connect using this same machine from the chat interface. However, when i try to connect to the server ffrom my laptop or any other machine i get a connection fail
The attached image shows 17 client connections. Unfortunately there are not 17 clients connected to that app. In fact, I add users to a userlist shared object in application.onConnect(), and that shared object only reports 6 connections. That is the accurate number of connected users.
Why the phantom connections? Even if I unload and reload the app, the 17 client connections return.
How can my userlist shared object and the client connection number be out of sync? This situation seems to be related in some way to an FMS memory problem that is cropping up all of a sudden. When I have larger numbers of connected clients - say 300, the client connections reported in the FMS console are ten times that - like 3,000 or more. Then the application is unloaded by FMS for being a memory hog.
The app itself is simple. It just stores each user in a user array and passes chat communication to and from these users. The last time this main.asc code was changed was 2/8/2010. No problems with it in the countless times we've used it since then. Why the sudden phantom FMS connections, memory hog behavior, and constant unloading of the app?
URL...I am running application instance that implements publisher/subscriber on video streams. Application code is very thin and can be removed completetly if required. In a specific time, randomly, disconnect messages stopped happening which causes the folowing:
1. Streaming service shows decrease of bw per second
I have FMS 4 and a videochat app (one-to-one), everything runs smoothly until about 750 users (it means about 1500 connections to fms), then I get a sudden drop in users connected to 300-400, no errors in the fms log and the cpu/memory usage is at about 25% just before this. I'm running Windows 2008 R2 Web (x64) and I've red somewhere that maybe this version of windows has some connection limit that could be the cause, but I've no idea of what could be happening.
i have FMS Installed And after its installed i also Install The Addon Auth to make a username and password so no one else can just start a Stream but after i install it and restart the server it wont work any the fms does work but the addon doesn't.
I have a FMS server running that I am hosting some VOD content on. I am using HTTP dynamic streaming to deliver this content. Everything runs fine from a technical standpoint, but I am unable to see the connection details in the Admin Console. I have loaded vod -> defaultinst under the applications tab in the console. No connection details populate.
I'd like to hear some opinions about using NetGroups VS. using direct p2p connections between everybody, considering there would either be 2,3 or 4 people max in the group.
I've now had two occurances of SWFVerification not enabling itself when the values are added to the application.xml for a client app.We've enabled it with an empty dir and any player will still get through.I tried to add a "not-really-the-player.swf" to confirm that it will reject all other connections, and they still get through.I've tried reloading the Application, and that also has had no effect.In a previous instance, only after the VHost is restarted, will SWFVerification start to work.This presents a rather large issue for us - as we host 100s of other client application connections on multiple boxes that do SWFverification, so it's not simple to just "reboot" these boxes.
Does FMS 3.5.2 in development impose restrictions via the origin of connections? That is, does it only allow localhost connections via ActionScript (NetConnection) or are external connections allowed?In other words, can I have FMS in dev mode on a dedicated server and then have my client code running on a different machine that opens connections to this development server?
In multiple instances, limiting the maximum number of connections for the same IP?You give me the answer is to limit a single instance, whether there is a way to limit the number of users of the entire connected to the FMS?
2 x Intel(R) Xeon(R) CPU L5410 @ 2.33GHz 16Gb RAM 2 Gbit ethernet channel OS - Linux CentOS 5.5 x86_64 FMS4 Interactive
Live stream parameters:
320Ñ…240 qua_=87 bw_=200000 kf_=5 fps_=18
So, after approximately 600 connections to one stream video becomes choppy (periodically freeze, slow motions). CPU-usage at this time is 100-120% (maximum is 800%, 100% per each core), network usage is 500 Mbit/sec.But second live stream(with few connections) at this server looks fine simultaneously with 600 connections at first stream.Experiments with recommendation in "Configuring performance features" documentation chapter(enabling/disabling aggregate messages and configure the size of stream chunks) do not help.
I've been trying to figure out FMS for about a week now. Because of company firewalls, all traffic must go through either port 80 or 443 (I configured this in FMS.ini). I also created a self-signed SSL certificate and configured the adaptor.xml file to use it. Nevertheless, here are the problems I am experiencing: For RTMPT & RTMPTE, I can't start more than 9 connections. Beforing configuring FMS to also use port 443, I could only get about 4 connections. Now I would like to say that I testing this between only two computers, the FMS server with apache web server and a client.why it will not at least start 10 connections which I know is the limit for the dev version? Also, when using RTMPS I can only start 1 connection. Why is that?
I have a web app that connects two computers and streams video to each other like a video conference. It currently works and also records the video in .f4v format. I am trying to figure out what I need to do to enable the client pc's to stream in H.264. Is it true that I have to use some type of encoder (like Live Encoder) to use H.264? If so, does Live Encoder have to be installed on the client computers? If that is true, is there a way to install the live encoder through our web app, configure it, and close it down when the conference is over, with no interaction from the client? Is there some type of plug-in that we could use that would eliminate the need for installing live encoder on the client computers?
It stands to reason that an OS would have a memory limit for the FMS process. I have a video chat app to which I may need to connect 300,000 or more concurrent users. Has anyone done this? Would this break the bank? What's the most you've ever had connected at once?
The video portion of my app goes through a good CDN, so I don't think there's a problem there. But my chat room is all on one FMS instance. I see something in FMS docs in "configuring performance features" that warns, "Adobe recommends that you ensure that your total system memory usage does not exceed the process limit of your OS". Granted it's in the context of configuring media cache, but wouldn't the same apply to chat room size if the numbers get high enough?
I know that we are an entire version behind the current, but this is due to problems with a failed upgrade due to a bug in FMS with streaming our (3TB) collection of .mp4's with the new version... that being said, onto our current trouble...
I am involved in project managment for a large video conferncing project involving over 100k simultanous connection and most likely to be increased 4 times by next six months.While figuring out the scalability area ,it was concluded that smart selection of hardware and origin-edge clustering will be suited most. Well while studying this documentation says it supports thousands of calls but unfortunately I could not determine the actual numbers. Can you tell me how many edge server will be needed for 100k simultanous calls/number of sessions per edge server and will we need license (4500$) for each edge server?
How long would it take for a raw beginner to learn Action Script 3 coding so as to have a basic understanding and an ability to code Flash.Also to be able to take existing code and adapt it to suit ones own flash project ?Are there any training videos on such or good books to read ?What is the best way of learning AS3 ?Is it in fact best to forget AS3 and get tuition in programming or some other discipline then take a look at AS3 and what timescale would be suitable to achieve this so as to meet the goal stated above
I have just started using Adobe Flash Media Server on Amazon Web Services. Does anyone know how can I monitor server statistics such as the number of active connections? I have connected using SSH to the server but I cannot know what to do from this point on?
I'm trying to find in the documentation why Apache is required for the feature set that FMIS 4.5 provides. Is it that there are modules/scripts/other built in to that Apache build that make the features work or is it that Apache is simply pre-configured to work and provide these services?
If it's merely a configuration issue, is there an outline somewhere of the set up so I can mirror the setup with iis?
My rtmp client connects to the server ok and rtmp data flowing ok for about 4hours, in the wireshark trace I can see that the server just stops sending data roughly after 4hours. The server version is 3.5.1, and I also try it on a 4.0 server the data stops coming from the server as well after 5+hours. There are no errors or any indication of session disconneciton in the diagnostic logs. I have also looked thru the settings in applicaiton.xml and server.xml I just don't see any obviious reason for the issue. I do send rtmp Acknowledgement to server after the client receives "window size" bytes, and the client sends ping response to server.
