I want to share some other discovery I made today:
In the last few days I was playing around with several other settings like --queue --expiry and --scan-time. And the more time passed, the less blocks my miners picked up. I tried all kinds of things to get it working again like resetting router, restarting rigs, ... I was about to give up when it occurred to me that the part of these settings might still have been stored and corrupted my mining settings. Then I remembered something you do when testing out new settings, to delete the "scrypt130511Tahitiglg2tc8192w256l4.bin" file in the cgminer folder.
No more than 10 minutes passed after deleting this file and my miners found already the first block again.
To make a long story short - if you play around with several settings and experience bad luck, try deleting this file - as this will reset and rebuild your proper mining settings.
Also, another tip:
If you add at least 2 --fallover-only -o stratum+tcp://pool.com:xxxx -u user -p password backup pools to your .bat file, you will increase your chances of finding a block while solo mining, as you will use the stratum's servers data to update your miner when a block gets found, so you are always working on the latest block. (And thus avoiding orphaned blocks.)
it's good practice to always start your .bat with del *.bin
.. how do you know your change in settings made any difference in such a short time frame? You could have been just unlucky at that point rather then it was the settings. I started my own testing myself, will report back how it goes .. 4 blocks found so far in 7 hrs .. i guess i'm pretty lucky tho tonight hehe .. and thanks for the tip about the failover, added it to my bat aswell
How I know that my changes made a difference in such a short time frame?
- Well I played around with the expiry and scan-time values, which made it much worse (no blocks were found in 24 hours), because if you are setting it too low, you are decreasing your chances
By setting each of these to 1 and 1, you're going for the insanity wolf level of difficulty: Every 1 second, go see if there's new work, and if you can't verify the work for 1 second, stop scanning.
Sensible settings are x0.25 - x0.5 your block confirmation time for expiry, and -1 for scan time: That ensure that you check for work on average 2-4 times before you actually expect to see new work on average, even though with stratum you don't need to, and you'll keep looking even in the case that the current work couldn't be verified to be current.
And then I just took off those three arguments, but did not delete my bin file. Obviously, my miners did not revert back to the standard cgminer settings. But they kept mining with the last expiry and scan-time values set. But as soon as I figured out that I need to reset my cgminer by deleting the .bat file, I found already 3 blocks in the last 6 hours.
That also shows you have to research more before you start playing around with those arguments (expiry, scan-time & queue). That's why I quoted that paragraph above, so if you want to adjust them yourself.
Also, add as many --failover-only pools (obviously the biggest ones you can find) as it connects you to their stratum data stream and updates you the second a block was found on the network (instead of the regular 60 seconds). I have right now 9 backup pools configured
.
