All network encoders disappear at random times...

[Kirby]

I'm running 5.0.2 with 3 network encoders (2 R5000 units, 1 cable box via firewire). All three are on the Sage server, running on separate ports obviously. Last night, all three disappeared from sage completely. I had thought it was an issue with the R5000 network encoders, but when I noticed in the log file that the firewire cable box went as well, I figured its something else.

So what can I do? Some of the log entries that might be worthwhile:

Sat 5/20 20:40:25.037 Seeker awoken
Sat 5/20 20:40:25.037 Seeker removing dead Encoder: Dish411-100227C on 127.0.0.1:5959
Sat 5/20 20:40:25.037 Seeker.endRecord(Sat 5/20 20:40:25.037) currRecord=FA[stationID=32645 IndyCar Racing time=Sat 5/20 19:00:00.000 dur=3:00:00]
Sat 5/20 20:40:25.037 VF.submitJob(VFJob[InactiveFile r=0.0 t=0 file=null ifn=F:\IndyCarRacing-Indianapolis500QualifyingPoleDay-1942157-1.mpg])
Sat 5/20 20:40:25.037 VF processing job VFJob[InactiveFile r=0.0 t=0 file=null ifn=F:\IndyCarRacing-Indianapolis500QualifyingPoleDay-1942157-1.mpg] nPlayin=false
Sat 5/20 20:40:25.037 VF discarding job VFJob[InactiveFile r=0.0 t=0 file=null ifn=F:\IndyCarRacing-Indianapolis500QualifyingPoleDay-1942157-1.mpg]
Sat 5/20 20:40:25.037 (Retryable)Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:40:25.037 VF processing job null nPlayin=false
Sat 5/20 20:40:25.037 VF thread is now waiting for 0:00:00.000
Sat 5/20 20:40:25.037 Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:40:25.052 Seeker removing dead Encoder: DCT2000-100178C on 127.0.0.1:6565
Sat 5/20 20:40:25.052 Seeker removing dead Encoder: DCT6200 on 127.0.0.1:6969

So am I looking at a Sage issue, and OS issue, or a JAVA issue? Heck I guess it could even be a hardware issue at this point? I can attach the whole log file or send it to anyone that might be wanting to see what else is going on, but its over 7mb.

I have also just noticed about 3 minutes prior to those entries above, I had:

Sat 5/20 20:37:46.654 EPG needs an update in 1134 minutes
Sat 5/20 20:37:46.655 EPG's works is done. Waiting...
Sat 5/20 20:37:51.802 UPnP ERROR:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:37:51.804 UPnP ERROR:java.net.SocketException: No buffer space available (maximum connections reached?): connect

and

Sat 5/20 20:38:36.427 Scheduler awoken
Sat 5/20 20:38:36.443 (Retryable)Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:38:36.443 Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:38:36.443 (Retryable)Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:38:36.443 Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:38:36.443 (Retryable)Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect
Sat 5/20 20:38:36.443 Error communicating with encoding server:java.net.SocketException: No buffer space available (maximum connections reached?): connect

I think its going to come to someone needing to look at this log file, because after the last set of entries above, the network encoders were not found by the scheduler or seeker, then at 20:39:28.709 the seeker found the network encoders again and added them, but then the seeker awoke again a second later, found the network encoders again, and added them again! Something funky going on here. I'll try and zip up the last 5 minutes of the log in case anyone is willing to take a look at it.

[Kirby]

In case anyone actually feels compelled to try and help on this, last night/this morning I have now encountered 3 crashes of Sage. So I am wondering now if its something to do with the Java version, as is being discussed in the "5.0.2 Crashes" thread.

I'm watching via my Sage server machine, 5.0.2, java 1.5.06, and the rest of the specs are in my sig. Sage wasnt doing anything but sitting there, not recording, not watching, and it just died. if this is a java version issue, i wonder why its just starting right now. I didnt previously experience issues with this version during the 4.1.x beta process, and even with 5.0.2 everything was fine until this past friday.

[Opus4]

Originally Posted by Kirby

... everything was fine until this past friday.

It sounds like you mean that the current setup was working until then... if that's true, did anything recently change on the system? Since you said this just recently started, I would look at any software that was added/changed just before this started. If nothing did, then I would check memory and other hardware.

If you need a debug log for SageTV, see this FAQ.

- Andy

[Kirby]

Yes, until Friday I wasnt having this issue. The only issue I had before was my remote recording directories disappearing, actually any recording directory that used a UNC path, which you and I discussed, and changed to a non-UNC path (E:\ and F:\) and didnt have any more issues with that.

The only thing I have added to the system is an external DVD player, and AnyDVD software (3 weeks ago). I have windows auto updates off, and havent done any other hardware driver upgrades in that timeframe.

I do have a long debug file (7.7MB) that I could send but I dont have room on my website to post it for linkage in this thread...

As for checking hardware, I will try tomorrow to do some stress testing on it (Prime 95 for 24 hours or so).