STV video playback stopping randomly

[VCRUser]

I changed the Video Renderer to Default, so I'll see if that makes any difference, then work my way through the other available settings if it doesn't help. Maybe it's trying to tell me that DirectShow support is present but not strong on my video card - of course I'm not even using the server's video card when I'm on PlaceShifter or an extender, so it's all a bit confusing. I think there's also a system option to have 3D acceleration on or off (it's currently on) - that could be relevant.

[wnjj]

I’m not sure but I think the server renders the UI using the server graphics system and then sends “frames” to the placeshifter and HDx00 miniclients.

[VCRUser]

I've tried switching out the renderer, increasing the Java heap size, and running in service-mode and non-service mode. No matter what, I am getting frequent hangs. Looking at the logs, I see all sorts of things occurring at the time things are hanging. Here's an interesting chunk:

Thu 4/1 3:53:36.210 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 750 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]
Thu 4/1 3:53:36.338 [Scheduler@35da479d] # Airs=32214
Thu 4/1 3:53:36.960 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 1500 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]
Thu 4/1 3:53:37.710 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 2250 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]
Thu 4/1 3:53:38.460 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 3000 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]
Thu 4/1 3:53:39.211 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 3751 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]
Thu 4/1 3:53:39.961 [AWTThreadWatcher-b8ca3aa4a67c@1dff914] EventThread-b8ca3aa4a67c Hang Detected - hang time = 4501 UILocker=Thread[EventRouter-b8ca3aa4a67c,5,main]


Other times, as best as I can tell, trying to match the times shown in the logs to times when things seem to have hung, is that it's working on updating the program guide, checking the status of the plugin repository, cataloging my recordings, scraping metadata, and who knows what all else. I wish I knew better how to interpret the logs - all I can tell is that the system is quite busy all the time. Is there a way to tell it to put off such activities for times when nothing is being watched, or perhaps to give it particular times when it is allowed to do those things, which correspond to times when watching is unlikely?

This is very frustrating because it happens often. Sometimes the Placeshifter UI hangs for so long that it tosses me back to the main menu or even out of the UI entirely.

[trk2]

If you read through the prior, similar posts of this problem (there are others but it's a difficult problem to describe which makes it tough to search for), the problem seems to be network or hard drive related. Sage support had this to say on the issue:

This is some kind of network or disk I/O delay. If you are using mapped
drives or UNC paths; then it's most likely a network delay....otherwise it's
a hard disk delay. Both are usually indicative of some kind of underlying
problem in the network or disk.

[VCRUser]

Since this is a brand-new build, with brand-new drives, a disk-issue seems unlikely to me. I have also done some drive-checking, with no issues found. There are no UNC or mapped drives involved - all are local, and the issues occur regardless of which drive holds the file I'm trying to play.

Do you know of any applicable network diagnostics I could look into? Windows' built-in network diagnostics shows no issues. It's an ethernet connection, and Windows indicates it's running at a gigabit.

The one clue I've found is that when I ran Task Manager during a huge slowdown last night, it was showing my 4gb memory at more than 95% used. That may have been because I'd raised the Java heap in a trouble-shooting effort. I've since put it back at the default and increased the swapfile size (rebooted after doing those, so assume they have taken effect). Since I *just* did the reboot, I'm now seeing more reasonable memory usage, but it's too soon to tell whether there will be an improvement in the hangup problem. I don't mind adding memory if that will help, but would like to figure out first if doing so would be helpful. Task Manager is not showing anything else using anything close to the amount of memory that Sage is using (Firefox is using 205mb at the moment, with 2 instances of Sage (presumably service plus gui) using 1.2G combined. I'll shut down Firefox when I'm done here.

The only other stuff running are background Windows tasks, like anti-malware, and none of these is using very much ram.

[VCRUser]

After a couple of days, I think it's fixed. Here's what I did:

1. Reverted Java heap back to the default. When I had it at a high value the problem got way worse, not better. That, coupled with the high memory usage suggested that all my memory was being gobbled up by the heap, and it was hurting more than it helped.

2. Turned off the default scan import directories every 2 hours, so that I would scan manually as I added new imported files. Since I don't import a lot of files often, that's fine. However, I'm not sure that was necessary, since I ran a manual scan while watching a recording, and didn't experience any halts - so I might turn it back on.

3. Turned off animations. I'm not sure that was necessary either - how many resources could they really be using? But the interface seems snappier without the animations, so I think I'll leave them off.

4. Increased my swap file from the default. My guess is that this plus resetting the heap to default were the real remedies. I might still install some more memory, just because 4G seems a bit skimpy on a 64-bit system, and it will hold up to 16gb.