GenTarkin
commented
5 years ago Well, unfortunately, even tho the pool compiled fine and is now running ... my same config of earlier version of nomp migrated to your new version + the custom working older multihashing repo I found ... results in 15% rejects, I dont know where the bottleneck is =( ... if its the odd compile w/ pool or if its simply network related. But, back before I had to do all this ... I had about 1% rejects. I did have to update wallet daemon software , so the performance issue may be there too but doesnt seem like it. A public pool on this network w/ my miner sees about 2.5% reject, nowehere near 15% that Im experiencing ... so the network for this coin seems healthy enough. Ugh... this is discouraging. Do u know, would there be anything in the multi-hashing that would effect reject rate? (I dont know how much that effects share / work handling and new blocks) Watching returned shares and new blocks on nomp it seems like a good 10 seconds passes after blocknotify triggers and shares are accepted ... any shares returned by miner within those 10 seconds are rejected.
makertronic
Just wanted to let you know, I and it seems lots of other folks run into issue w/ npm update failing on multi hashing. I tried all sorts of 0.10 versions, none worked. After trial and error of downloading some of the other forks of nomp... some of them compiled (because their multihashing was older) and some didnt(for similar reasons as this repo). Ultimately my solution was having to fork stratum pool to my own repo, then redirecting the multihashing dependency in package.json to point to an older github fork that I found worked ... in this case foxer666's fork of multihashing compiled fine. After redirecting that dependency, your repo of nomp compiled fine. Pool fires up and works. I realize that newer algos may be missing because of the older multihashing ... BUT ... I dont use the newer algos so that doesnt effect me.
The errors I kept running into were in the cryptonight.o compilation during npm update. Perhaps recent changes to that source cause the problem? I dont know. But either way, just a heads up.