{{Quickfixn}} Heartbeat delays and re-start MDR

[Øyvind Sævareid Ellefsen]

Ladies, gents;

I have seen some issues in my heartbeat lately, I can see that 2-3 times during the day, I am sending a heartbeat with 35-36 seconds instead of the configured 30 seconds.

My question is; may that be caused by "too much" work for the application? Honestly, I do not believe such, the application consumes only 2-3% cpu, but still there a lot of operations going on for each receive of MarketData.

Is the heartbeats handled by it's own thread, or is it running through same thread as the IApplication?

What happens next, is that my broker sends me a TestReqId, and since I have already sent an heartbeat (delayed) it does not accept my answer on the testreqID (Incorrect TestReq). I am then logged out, but will be logged in again automatically. I have tried to put a startmarketdatarequest on logon, but it looks like I initiate this request to fast. Is there any best practices for how to handle such disconnects, and automatically start the marketdatarequest again?



Brgds
Oeyvind Ellefsen
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quickfixn.com/pipermail/quickfixn-quickfixn.com/attachments/20131206/626396de/attachment.htm>