[multicat-devel] Issues with recording and playback

[Eduardo Vieira]

It seems the aux is still going wrong with me. I've copied a part of
ingests_validade's output below. I can provide you the aux file if you
want.
Have you tested with the splitting file feature turned on ? What
kernel version are you using ? Maybe I can try to get the same here.

PS. I'm already using some file system tuning options. Here is my
fstab line for the recording partition:
/dev/sda4       /iptv   xfs
defaults,noatime,nodiratime,nosuid,nodev,allocsize=64m    0 0

The block size was set to 4096 when I formated it.

0: diff exceeded 1327438439525
0: diff too low 0
1: diff too low 0
2: diff too low 0
3: diff too low 0
4: diff too low 0
5: diff too low 0
6: diff too low 0
6: diff too low 0
7: diff too low 0
8: diff too low 0
9: diff too low 0
10: diff too low 0
...
60470: diff too low 0
60471: diff too low 0
60472: diff too low 0
60473: diff too low 0

2012/1/23 Christophe Massiot <cmassiot at openheadend.tv>:
> Le 23 janv. 2012 à 22:28, Eduardo Vieira a écrit :
>
>> I have a machine streaming a TS file (using VLC) to a multicast
>> address in the network. On the other side I have a machine recording
>> and playing back (when requested) to a setopbox, over unicast. The
>> recording is beeing made using directory out mode, with a rotate time
>> of 5 minutes.
>> The STB is connected to a TV, and the problem is easily perceived by
>> looking at the image, in a regular basis the video shows some macro
>> blocks or get stuck for a moment, depending on the case.
>> I'm not making any modifications on the other parts of the test
>> system, except the machine that is running multicat.
>
> OK, then can you confirm that the .aux file is *correct* (attached program may be helpful ; you may need to adjust the thresholds) ? That is, that there aren't a gap and then 15 packets in a row with less than 1 ms between them ?
>
> What I suspect is that at the moment stuff is flushed to disk, the playing multicat may have difficulties reading. If it is the case, blockdev --setra 4096 /dev/XXXX may help.
> _______________________________________________
> multicat-devel mailing list
> multicat-devel at videolan.org
> http://mailman.videolan.org/listinfo/multicat-devel
>