doWatchdog took too long

Wowza Streaming Engine
1

Hello,

We are facing an issue where all our encoders are disconnected from time to time. They reconnect automatically a few seconds later. We noticed in the logs that it happens when there is a trace mentionning “doWatchdog took too long”. After this line, we have 2 lines for each encoder (zero sized function and killConnection).
What is the meaning of this message about Watchdog ? What could be done to fix this problem ?

2024-06-04 17:33:36 CEST comment server INFO 200 - VHostWatchdog: run: doWatchdog took too long. Interrupting thread.defaultVHost - - - 732530.965 - - - - - - - - - - - - - - - - - - - - - - - - -
2024-06-04 17:33:36 CEST comment server ERROR 500 - Zero sized function (client:336856530:FMLE/3.0 (compatible; FMSc/1.0)): type:20 size:0: - - - 732531.179 - - - - - - - - - - - - - - - - - - - - - - - - -
2024-06-04 17:33:36 CEST comment server INFO 200 - RequestProcessData.killConnection: 336856530 - - - 732531.18 - - - - - - - - - - - - - - - - - - - - - - - - -

Thanks

2

Hi,

please open a support ticket from your Wowza account, so we can help in your individual use case. Also include your conf and log files please.

Any additional information you can give about the source will also be helpful.

The error indicates an issue with the inbound connection, but to correctly debug, we’'ll need the full logs and the WSE startup sequence.

Regards,
Bernd Backhaus
Wowza Support
WOWZA | The solution you start with, the partner you scale with.

Wowza Video is here! Look at all the new features that will help elevate your business.

Manage your Support cases online
Bookmark our FAQ page
Bookmark our Status page