There are a number of criteria for when completed WU's get reported (i.e., those in a 'Ready to Report' (RtR) stage - the data itself, generally gets uploaded straight away).

The three most common ones are;
1) When your computer is requesting more WU's (which is dependent upon your cache/the estimated amount of work remaining etc), it completes the reporting process for those that are in the RtR state,
2) If a WU has been in the RtR state for 24 hrs, it then completes the process off, or
3) If the WU is within 24 hrs of it's deadline, then it'll get reported straight away.

As I say, there are several other criteria for when/how WU's get sent back - and these (mainly) depend on your network settings (i.e., if your connection to the internet is set to download/upload WU's at a certain time etc.).

Gang.... thanks for the responses ! Now I understand how the reporting mechanism works (until I have another question) .

CJSL

AFAIK, sending the data back, just entails that - whilst reporting results, requires access to the database, thus more transactions are needed.

If a back-off timer is running or there is a ''connect every...'' setting, that could be the case. In your situation, re the other thread v.v. HFCC exclusive with limited supply that could happen more frequent. Whether it cannot be the issue here, we can speculate. Without long logs with extra flags that print out backoff times, work fetch scheduling etc etc, it's a guess. Just know the basic rules how most clients are supposed to work... much of it written up in Wikis and Start Here FAQ forum.

--//--