I'm seeing the "low difficulty share" issue here now. Took about 24h of testing for it to show up.
I saw this begin happening after a disconnect from extremepool at 11:30 EDT. Did this happen to others at the same time?
I have added a more complete restart mechanism to mask this even more, but would like to identify the underlying source of the failures.
Miner updated, patch here:
https://github.com/dave-andersen/boolberry/commit/06998dc0ed2f63683dadf5ea293c8217575c9b46
Will keep testing. C_Z, if you have any insights to how the scratchpad copy might get permanently desynchronized with the server, I think that's where the error lies. Hopefully this more complete restart will mask it, but re-downloading the scratchpad will increase bandwidth use, so it'd still be good to fix.
-Dave
This fix appears effective.
2014-Jun-15 14:08:52.666932 Failed to submit share! Updating job.
2014-Jun-15 14:08:52.666939 Too many submission failures. Something is very wrong.
2014-Jun-15 14:08:52.666944 Excessive failures. Sleeping 10 seconds and restarting...
2014-Jun-15 14:09:02.679750 Mining with 8 threads
I'll keep digging, however, because of the potential bandwidth use of having to re-download the scratchpad twice a day or so, as the blockchain gets larger.
