Using the FFmpegFrameGrabber.start() method, one thread is blocked。

[Aismy]

You can see that one thread is blocked in the avformat_open_input method after entering the start() method, causing other threads to be unable to acquire the lock to process all the BLOCKED states.

"media-87" #217 prio=5 os_prio=0 cpu=2532193.98ms elapsed=3003.56s tid=0x00007fd42c003000 nid=0x366c08 runnable [0x00007fd21c66e000] java.lang.Thread.State: RUNNABLE at org.bytedeco.ffmpeg.global.avformat.avformat_open_input(Native Method) at org.bytedeco.javacv.FFmpegFrameGrabber.startUnsafe(FFmpegFrameGrabber.java:972) - locked <0x00000005becf30f8> (a org.bytedeco.javacv.FFmpegFrameGrabber) at org.bytedeco.javacv.FFmpegFrameGrabber.start(FFmpegFrameGrabber.java:903) - locked <0x00000000807230b0> (a java.lang.Class for org.bytedeco.ffmpeg.global.avcodec) at org.bytedeco.javacv.FFmpegFrameGrabber.start(FFmpegFrameGrabber.java:898) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:750)

"media-92" #222 prio=5 os_prio=0 cpu=68.47ms elapsed=2963.55s tid=0x00007fd320003000 nid=0x36762f waiting for monitor entry [0x00007fd21c16e000] java.lang.Thread.State: BLOCKED (on object monitor) at org.bytedeco.javacv.FFmpegFrameGrabber.start(FFmpegFrameGrabber.java:902) - waiting to lock <0x00000000807230b0> (a java.lang.Class for org.bytedeco.ffmpeg.global.avcodec) at org.bytedeco.javacv.FFmpegFrameGrabber.start(FFmpegFrameGrabber.java:898) at reactor.core.publisher.LambdaMonoSubscriber.onNext(LambdaMonoSubscriber.java:137) at reactor.core.publisher.FluxPeek$PeekSubscriber.onNext(FluxPeek.java:192) at reactor.core.publisher.FluxSubscribeOnValue$ScheduledScalar.run(FluxSubscribeOnValue.java:178) at reactor.core.scheduler.SchedulerTask.call(SchedulerTask.java:50) at reactor.core.scheduler.SchedulerTask.call(SchedulerTask.java:27) at java.util.concurrent.FutureTask.run(FutureTask.java:266) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$201(ScheduledThreadPoolExecutor.java:180) at java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:293) at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) at java.lang.Thread.run(Thread.java:750)

this is my code： FFmpegFrameGrabber grabber = new FFmpegFrameGrabber("http://xxxxx"); grabber.start();

[saudet]

Please try to call startUnsafe() instead of start().

[Aismy]

startUnsafe() will also acquire the lock ah, my method is blocked in the startUnsafe() method inside ah, and blocked for several days!

[Aismy]

Now blocking at avformat_open_input() in startUnsafe()

[saudet]

Make sure to set a timeout value