Opened 12 years ago

Closed 12 years ago

#2480 closed defect (fixed)

sid randomly gets blocked (for about 70 secs) in trunk, works fine in 1.10 (5741) stable

Reported by: spock Owned by:
Priority: minor Component: General
Severity: medium Keywords:
Cc: Sensitive: no

Description

Revision

6593

Issue Description

When watching through an otherwise stable cccam reader. It will usually start with a high ecm (700-1200ms) that gets rejected, and then oscam will continue rejecting the sid for just over a minute before picture returns. Restarting oscam as soon as picture freezes will immediately bring back the picture.

This does not happen with stable 1.10/5741. Here the screen will only show at most some pixelation due to a slightly too high ecm time. But playback continues without any other issue.

When the issue occurs

Almost completely at random..

How the issue is reproducable

Don't know of a way to trigger it manually, it's too random.

See logfile, -d65535, interesting lines around line 25930 and forward.

Attachments (2)

oscam.log (1.7 MB ) - added by spock 12 years ago.
2-oscam.log (198.4 KB ) - added by spock 12 years ago.
2-oscam.log

Download all attachments as: .zip

Change History (9)

by spock, 12 years ago

Attachment: oscam.log added

comment:1 by szili_boy, 12 years ago

I tottaly agree with this. Stable works fine.

comment:2 by Deas, 12 years ago

and where is the log from the server? this client log is quite useless as it only is the half story...

comment:3 by spock, 12 years ago

Currently not possible to get that.

Anyhow if I use CCcam as a client, I might get one lost ecm (one freeze), then it returns to normal quickly. With newer oscam I have to wait 30-90 secs or sometimes even before it recovers. Unless I restart oscam, in which case picture decoding returns immediately.

Bug or by design? Just close the ticket as invalid if you think this is normal behaviour.

Last edited 12 years ago by spock (previous) (diff)

comment:4 by Admin, 12 years ago

Seems to me like it's the loadbalancer which blocks this. Did you try without it? Are the loadbalancer settings the same (look in WebIf; I'm not remembering this exactly but it might be that some default values changed her)

comment:5 by spock, 12 years ago

The problem remains with loadbalancer set to either 0, 1 and 10.

Does this seem related? http://www.streamboard.tv/wbb2/thread.php?postid=415992

That's tickets #2460 and #2479

Last edited 12 years ago by spock (previous) (diff)

by spock, 12 years ago

Attachment: 2-oscam.log added

2-oscam.log

comment:6 by spock, 12 years ago

Just attached another log of the same problem, with loadbalancer disabled.

In this log problem starts after line ~1443. Decoding just stops and doesn't return until I close the stream and start it over again, which you can see after line 2704 or there about.

This is a recurring problem that happens quite often, on many providers (although much less on some providers).

Tried every possible configuration to fix it, but no luck. Always have to restart oscam or the stream/zap (although sometimes a restart of oscam is necessary). I should also note that I have no particularly fancy or special setup, just using oscam with dvbapi together with usually 3 locals and two proxies/remotes. Locals are never a problem.

The proxies/readers obviously do some hickups sometimes, be it a setup problem on the remote end or my/their Internet connection. Still older oscam and CCcam can recover from it quickly, but newer oscam gives up more or less for good (until restarted).

Last edited 12 years ago by spock (previous) (diff)
Note: See TracTickets for help on using tickets.