Media Server :: FMSEdge.exe Runaway Process On 3.5.2 / Win2003 Server?
Jan 14, 2010
Default install of Flash Media Server 3.5.2 r654 on Windows 2003 Server Sp2. 16 GB of RAM. Just testing with the out of box Flash application, and simple 'hello world' rtmp streaming video app, we see several times a day at seemingly random intervals FMSEdge.exe will jump to 100% and peg the CPU and stay there. Until it is killed, the Flash applications are unresponsive until it is killed. As one would expect, this is a showstopper for any serious use -- Keep in mind this is an OOB install. There was a previous thread mentioning a similar issue (ie FMSEdge pegs cpu) coinciding with the virus scan, however no one responded to solve that issue.
I have a problem with setting up my FMS and to NICs
NIC #1: public IP NIC #2: private IP
On #1 I will setup port 1935 and 80 - Clients can connect to my livestreams
On #2 I will setup port 1935 and RDV for the FlashLiveEncoder and the RemoteDesktop
My problem is, when I activate the NIC #2 the FMS is noct reachable from the public NIC.I think it is a FW problem or a problem withe application.xml. how to setup my server (NIC, application.xml and FW) to run how I like?
I'm a systems administrator at an ISP running many FMS 3.5.1 instances for an array of clients. About half of these FMS 3.5.1s fall over dead after 2-3 weeks of runtime. Specifically, 'fmsedge' becomes unresponsive to http/rtmp tests on port 80; NOT on port 1935. The process appears to be doing work (cpu time is nonzero), and it answers the phone - I can connect to port 80. But I can't pass traffic to the daemon.
Logs give no indication anything's wrong. I can't strace or truss java with any legibility, due to fmsedge living atop a multithreaded virtualmachine.This occurs regularly, but does not occur on /every machine/; only about half of them. Where it happens, it's at 2-3 weeks after process launch.how to collect additional diagnostics, so that I can find what's breaking in fmsedge and either prevent it.I am running on CentOS 5.3 (like RHEL-5), 64-bit; I install nothing but Adobe FMS, and used Adobe's installer.
I configured my FMS for listening on port 80 and redirect non-FMS requests to apache on port 8080 in the same server machine. It works well, but I've noticed several tcp connections from fmsedge to apache in Established status (it is simple to see through "netstat -nap | grep 8080" command) that stay forever (until FMS stop or restart). With a sniffer capture I seen that sometime the fmsedge requests to apache a tcp connection but it (fmsedge) doesn't use this connection. Then the apache server send a syn,ack tcp request and always fmsedge send a dup ack tcp response. So, the connection is active but not used, then apache doesn't close it.The kind of requests for apache are http get to js or php files. I don't understand why fmsedge create these connections.
How and when these errors happen and how to recover from these errors. Does it write additional info to linux system logs.fmsedge related error seems to be when it reached 2.6g it is becomes unresponsive.
I am wondering what values should I put in the Process tag of Application.xml considering my application. I've done some studies but I am still unsure and would like some expert opinion To help you help me : I have a video chat application that comes with a contact list, so think skype on the web.
[Code]...
I set a RollerOver because sometimes RTMFP crashes on my server, and by putting the RollOver it seems to help, by re-newing the core process. However when clients are connected before the rollover, then they will not see their friends going online after the rollover which sucks.
I am sometimes (sporadically) facing the problem, that I cannot connect to FMS server with FMLE, using username and password.When I then restart FMS everything is working correct again. A customer discribed that he was not able to reconnect to the fms, as soon as he closed the FMLE without stopping encoding process?I am using the default authentication plug-in in FMS 3.5.6 on a 64-bit Debian.Everything else is working correct, only reconnecting with encoding-software (like FMLE) is not working.Is there any possibility to get the authentication working again, without restarting the whole fms process?
I am experiencing some failures with FMS 3.5.1. Searched the logs and it appears the following error repeats itself more than 20 times per second in the admin logs: Failed to wait for process condition: errno(22).
The line: Master process not found, or multiple masters
appears in /var/log/messages of a Centos server running FMS 3.5 right around the time Flash videos stop working on a site we built. Oddly, nothing appears in /opt/adobe/fms /logs/ master.xx.log at the corresponding time*. Is there likely to be any more illuminating information somewhere on the server, to help us identify what caused Flash to stop working?
I've been experimenting with Dynamic Streaming while in the process of writing a tutorial. The documentation is quite clear on a few points: The keyframe interval in the various encodings should be shortThe bufferlength should be at least 2x the keyframe intervalThe player should sense a bandwidth change, by default, within the 4-second sampling interval and call for a switch. Then, the switch could take as long as 2x the keyframe interval after that. What I'm finding is wildly different behavior than this. It takes anywhere from 10-15 seconds for the player to notice the change and call for a switch, then another 20-40 seconds for the switch to happen. When switching up to a higher bitrate stream, this just means the user gets low bitrate video for longer than they ought to. But when switching down due to falling bandwidth, the buffer runs out and the user stares at the rebuffering sign for a lengthy time - long enough to give up on watching the video, for sure.
I've encoded an H.264 MP4 file at 64, 384, and 768 kbps, at 30fps and an "every 60 frames" keyframe interval. I've streamed it rtmp via two different CDNs that use FMS 3.5, into two different Flash video players (JW Player and Flowplayer). I've restricted my bandwidth on Windows XP with Netlimiter 2.0; and on the Mac with 'ipfw'. I've set bufferlength between 4 and 10 seconds. I've tested switching up and switching down. For up, I start with a 200kbps bandwidth limit. The video starts OK with the correct stream, then at 5 seconds I open up the bandwidth to unrestricted. For testing down, I do the opposite: start at unrestricted and then at 00:05 restrict to 200kbps.
My test page, with both players and sample code is at [URL] n-flash-bitrate-switching/ I also have a couple of screen recordings there showing the behavior of the whole process, both switching up and switching down. I thought I've done everything right here - paid attention to every documented detail, but it works rather poorly. Can someone explain whether this is expected behavior, if the players have implemented dynamic switching poorly, or if I'm doing something wrong?
I came across a very strange issue. There was a session in which no fso has been recorded. It was a session of around 110min with more then 10 users as per the logs. Flv have been recorded for the session and I can see the chat text file which has been written from server side script but no fso has been recorded for the whole session. While cross checking the logs access log and Master log i saw that the last message recorded on that particular session was around 14:47 and in master log i see following lines:
2010-04-18 14:52:56 52768 (i)2581223 Core (56012) is no longer active. -2010-04-18 14:52:56 52768 (w)2581256 Core (56012) _defaultRoot_:_defaultVHost_:::_0 experienced 1 failure[s]! 56012 is the core process on which this session was running. So is it somehow related? I mean no fso being recorded on server with core recycling? In access log these are the last 2 messages on this core:disconnect session 2010-04-18 14:52:20 Central Daylight Time 186.24.19.5 56012connect session 2010-04-18 14:52:46 Central Daylight Time 186.24.19.5 56012 FMS server version is:FMS is 3.0.2Windows server 2003.
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'm building a video conferencing applicaition for a portal. But now when considering which version of flash media server to buy, I run into some problem. So can anyone helps me about comparing this two versions: Flash Media Streaming Server and Flash Media Interactive server, like if I use Flash Media Streaming Server, some function like NetStream.pause() may not work or something like that? I'm not sure if this is a foolish question but please let me know.
The company I work for installed FMS on a server running Centos for a customer who needed streaming media capabilities. It worked fine for 3 months until today the customer called us and informed us that their client program can no longer connect to the streaming media server. No one has done anything to the server to cause this to happen. When attempting to access the admin console to diagnose the program, the streaming server returned a 500 Internal Server Error. So I tried another page. Same result. Reboots of both the hardware server and the FMS have changed nothing.
One server - is streaming server with Flash Media Interactive Server 3.5that host the application with asc files in FMS application directory.Second server - is the IIS web server that host thehtml,aspx,swf etc.. files.So basically i heve swf file on one server that have to connect to live streaming via rtmp on different server with different ip address.I did not find any clear explanation on this crossdomain rmtp issue.ys only HTTP can use crossdomain.xml policy filebut not rmtp.So what policy security procedure need to be done in order to enableswf file hosted on one server to connect and show rmtp live cam broadcasting from another server?
Can Adobe Flash Media Streaming Server 3.5 run on AMD Athlon Dual Sock Quad Core?I just requested a Dell server to be added to our farm to run as a Media Server and to my surprise, while reading the requirements for FMS it states the following: 3.2GHz Intel® Pentium® 4 processor (dual Intel Xeon® or faster recommended)
I'm trying to troubleshoot a Flash Media Server working with a little video playback application I wrote a few years ago that has suddenly stopped working.I'm using CS3/Actionscript 3.My app uses the FLVPlayback Component, and was working well last time I checked. I recevied a report that the videos stopped working, and have been looking into it.I figured I'd add a bunch of event listeners to the FLVPlayback's ncMgr.netConnection so I could get debug info on things like io errors, net status, etc.The problem I'm running into is that the netConnection is null when I set it to anything on my Flash Media Server, and adding any event listeners to this netConnection throws errors.Here's what I've tried so far:
Playback of a local FLV file works fine.In the FLVPlayback documentation, I found an example and stole the URL of the stream they were using in the example, and that works fine, although it is an HTTP protocol stream rather than RTMP.Any attempt to access FLV files on my Media Server, which has worked fine in the past, basically cause the FLVPlayback object to sit and hang in "buffering" mode and never progresses beyond this point.The netConnection object in this case is null.Here's my code:[code]........
Again the purpose of this is to troubleshoot the video streaming from the Flash Media Server, and it seems like there is no netconnection to the server being created.Does this mean that the server is not working, or is there a problem with the way I'm trying to access the content on the server?This was all working fine before, and I have set up the server-side application .ASC files and such to allow things to work fine on the server end.
I'm trying to make a software which sends video and audio data to a flash media server by using RTMP protocol. Currently, my program can communicate with a flash media server correctly. RTMP specifications does not describe about the raw data in video/audio messages, so I muxed raw H.264 and AAC data into video/audio messages and sent to the server. The server seems to accept them, but a video player cannot playback the stream sending from the server. The player just says "Loading..." For a test purpose, I sniffed the network packets between Wirecast and the flash media server and ripped off only video and audio data. Then, I muxed those data into video/audio message and sent to the flash media server. In this case, the video player connected to the server can playback the stream correctly.
I checked the stream sent from Wirecast, the stream seems not to be H.264 raw data because those data are not started from 0x17 instead of H.264 start code. With those situation, I am wondering what kind of container format I should use for H.264/AAC data to the flash media server.
The instructions for installation only cover installation on "localhost". I'm wondering if I could use the development server for a small scale website depolyment with restricted traffic, from a commercial server, such as goDaddy?And if so, what''s the method? Do I download the package from Adobe to my local hard drive and install over the network, or copy the package to the server and install it online?
The $995.00 version is listed as not supporting server side playlists. Is it possible to implement internet TV station like functionality using that version? I would like to have a list of files to play out and have it so those connected are all viewing the same thing.
I have to a problem using the Flash Media Interactive Server Feature Explorer. I want use the sample: RecordStream. I can see the instance "RecordStream" in console FMS 3.5. and show me the video in app AIR, but does not save the .FLV in my server.
I have Flash Media Streaming Server 3.5 (not Interactive) running on RHEL5.5 x86_64 Linux.All is working well, however how do I prevent unauthorized access to connecting to the live stream and streaming content?How can I setup the server to require a user and password to stream live media to the server?I am new to this product and I have been reading some documentation but I have not found a clear cut answer on how to force a username and password to connect to the server to stream live content only.I am using the Adobe FMS Apache install, what files need changing?[code]I want to lock down a person from connecting to the server on the public internet and starting a live stream?Can this be done with a user name and password?
We purchased FMIS and we are encoding large 15+ hour MP4 recordings using flash media encoder. When opening these large files for playback, which have not been opened recently the player displays the loading indicator for up to 4 minutes! Once it has apparently been cached on the server it opens immediately from any browser even after clearing local browser cache. So a few questions for the experts
1. Why is it taking so long to load the file. Is it because the MP4 metadata is in the wrong format and the file is so huge? I read somewhere that Media Encoder records with incorrect MP4 metadata is that still the case?
2. Once its cached on the server, exactly how much of it is cached. Some of these files are larger than 500mb.
3. What fms settings do you suggest I change. FMIS is running on windows server R2 64 bit, but FMIS itself is 32 bit. We have not upgraded to the 64 bit version. We have 8GB of ram. Is it OK to set FMS cache to 3GB. And would that only have enough room for 3-4 large files, because we have hundreds of them.
