Idea number 2, let the server dish out 3 or 4 of each wu then only the first 2 need to be verified and the other 1 or 2 get server aborted, this will keep everybodys stash of wu`s low, No bunkering.
Im not going to use my Note3 this time lol
I'm not a fan of this idea as resources such as bandwidth gets wasted. Some networks can't be left active at all times and so limited times for activity is used. I know that one of my teams big hitters had connection reliability issues and there is a bug in BOINC that will error out all work units when the connection goes south. It has been there for years and nothing has really been done to fix it. So, to guarantee enough work between outtages requires a cache.
And to be clear, when you are caching up enough work with several 4P rigs that have 60+ cores/threads in them, it can take a lot of work to restock small caches often. Large caches means not spending several hours every day stocking up...
Last edited: