Yup up again.

CSC causing > 50% rejects... Guess the coin is too fast.  May need to Downgrade this coin on the profitability list when the Diff is low but increase when diff is higher.

Looks like there might be some problems recently, rounds 1608-1610 ?

A 22 minute CGB round with 0 blocks.
A CSC round with a time of 00:00:-1
A 2 minute MEC round with no blocks, and now we're on MEC currently for a long time.

CGB having PoS problems. All orphans.

CGB has been temporarily disabled until it's fixed.

We are looking into some issues with the last few rounds...will update here in a few!

CGB has been addressed (for now).  the 1611 MEC round stats aren't correct, we are working on updating them, we earned ~1500 coins or so.

MEC round cleared up

Some updates....

CAP and CGB currently disabled.  We've got a new CAP client, though I ran into an error or two running a script, haven't had time to confirm if it was just an older artifact or an actual daemon problem. Hopefully we can re-enable tomorrow.  Undecided on what to do with CGB currently.   If its expected we may run into 40-50 blocks at a time being orphaned (though we've never seen it until tonight), then it likely deserves a fairly decent profitability hit.  I'll dig into this more tomorrow/later as well.

We've also been adding some LTC traded coins, such as FLO, MEM, XNC, IFC.  This is where the bulk of our time has been spent recently, ensuring the trade flow for these coins is smooth and functional.  These are currently in a somewhat 'beta' status, we are keeping a close eye on their performance, and will look for more to add if they prove profitable.

On the short term agenda:

* Code fix for minor daily outage
* Change ledger to be total historical record not just last 50
* Roll out diff caching to Hybrids (its been in place for non-hybrids for a week now)
* Roll out optional static and static-minimum diff options, likely including a drop in pool target diff, to help our low end miners hit their cached diff quicker.
* Start looking into implementation of 2nd stratum server, ideally on the North American West Coast.

More or less in that order.

I agree on the intensity and thread concurrency settings, those are the most important to max out hashrate. You also might look into the number of threads. (  -g 1 or -g 2 )
Also important : the catalyst driver. My 6990 does best on 11.3, 7970 however needs 12.6

Note : I stopped using --auto-fan for temp control. This is just killing fans. Had 3 died because of this.
Rather use CGwather, as it lowers intensity when temp is too high.

Another word of advise:

THE most important cgminer flag is:

--verbose

Without it the miner's not telling you if you are getting any errors at all. So while your KH/s may be sky-high (like the poster's above us), you're just mostly spewing garbage to the pool without even realizing it.