Hello, I have a different setup than most, I have 25 ssds in a server. That server is maxed out so I’ve been farming them remotely through my 10G network. Recently I’ve stopped signing hashes due to the time limit. After checking sever other things I started testing all the releases starting with the newest. Finally when I got to the 18th part 2 it started working again. Now from what I’ve gathered through all my testing is that mostly likely something has changed since the 18-2 release that has effected my farming. Does anyone know what this change could be? Could we change it back so future releases support farming over a network?
Farming during initial plotting was enabled by default on large CPUs in dec-20 and significantly improvements to cache handling were done in dec-22. Assuming you have plotted, I see no reason why it wouldn’t farm.
Since dec-20 doesn’t work for you, it might be that you’re still plotting some sectors and that is causing issues, then you can specify --farm-during-initial-plotting false to change it to the old behavior.
Thank you for your reply. Not all the drives are plotted, some are plotting and some are replotting. Farming while plotting works with 18-2. I’m not sure why it wouldn’t work with the newest. I’m going to try the newest again and see if I can nail it down with some thread commands or something. I’ll report back if I get it working.
I’d try dec-20 first to narrow down a set of changes which might have caused this, but either way I see nothing obvious that would be a reason for a performance regression in recent releases.
Just wanted to confirm you noticed that the flag --farm-during-initial-plotting has been updated to be a boolean value so the equivalent is now --farm-during-initial-plotting true.