Here is an example from log files
We thought a block was found but it was rejected by the daemon, share data: {"job":"d3a2","ip":"::ffff:x.x.x.x","port":2053,"worker":"t1xxxxxxxxxeL","height":236147,"difficulty":0.2,"shareDiff":"4596.03419805","blockDiff":1351.176112912,"blockDiffActual":1351.176112912,"blockHash":"0000007212f19755fb017650d29e3823cf44a2b5aca1cf90baa36f3152951c0f","error":{"unknown":"check coin daemon logs"}}
ERROR: CheckEquihashSolution(): invalid solution
ERROR: CheckBlockHeader(): Equihash solution invalid
Have you considered that the user might just have used the wrong hash algorithm param when connecting to the pool?
If the protocol is different it goes right away as an error on the first submit.
I have tested all the pools listed on the first page. They all report shares mined with a wrong algorithm as correct back to the miner:
$ ./optiminer-equihash -a equihash200_9 -s zer-eu.forgetop.com:2053 -u <...> -p x
[2018-01-17 01:42:59.055] [default] [info] Optiminer/Equihash 2.1.2 (C) Optiminer 2017
[2018-01-17 01:42:59.055] [default] [info] Connecting to zer-eu.forgetop.com:2053.
[2018-01-17 01:42:59.149] [default] [info] Using 2 verifier threads.
[2018-01-17 01:42:59.149] [default] [info] Using highly optimized kernel code for Fiji devices.
[2018-01-17 01:42:59.149] [default] [info] Reading bianry kernel from bin-223600/Equihash200_9-Fiji.bin
[2018-01-17 01:42:59.150] [default] [info] Autodetected '--intensity 5' for device 0.
[2018-01-17 01:42:59.192] [default] [info] Extranonce is '60000043'.
[2018-01-17 01:42:59.233] [default] [info] Mining target is 0050000000000000000000000000000000000000000000000000000000000000
[2018-01-17 01:42:59.234] [default] [info] Got new work.
[2018-01-17 01:42:59.260] [default] [info] [GPU0] Device info: {"id": "0/0" "name": "Fiji" "vendor": "AMD" "driver": "2482.3"}
[2018-01-17 01:42:59.314] [default] [info] Connected to zer-eu.forgetop.com:2053.
[2018-01-17 01:42:59.464] [default] [info] Using highly optimized kernel code for Ellesmere devices.
[2018-01-17 01:42:59.464] [default] [info] Reading bianry kernel from bin-223600/Equihash200_9-Ellesmere.bin
[2018-01-17 01:42:59.465] [default] [info] Autodetected '--intensity 5' for device 1.
[2018-01-17 01:42:59.570] [default] [info] [GPU1] Device info: {"id": "0/1" "name": "Ellesmere" "vendor": "AMD" "driver": "2482.3"}
[2018-01-17 01:42:59.773] [default] [info] Using highly optimized kernel code for Ellesmere devices.
[2018-01-17 01:42:59.773] [default] [info] Reading bianry kernel from bin-223600/Equihash200_9-Ellesmere.bin
[2018-01-17 01:42:59.773] [default] [info] Autodetected '--intensity 5' for device 2.
[2018-01-17 01:42:59.879] [default] [info] [GPU2] Device info: {"id": "0/2" "name": "Ellesmere" "vendor": "AMD" "driver": "2482.3"}
[2018-01-17 01:43:01.015] [default] [info] [GPU2] Share accepted! (1 / 1)
[2018-01-17 01:43:02.549] [default] [info] [GPU0] Share accepted! (2 / 2)
[2018-01-17 01:43:03.061] [default] [info] [GPU2] Share accepted! (3 / 3)
It seems the pools internally filter those incorrect shares which is good. But they should be reported back to the miner as invalid.
The pool always get warnings when zcash solution is submited:
Share rejected: {"job":"ccd0","ip":"::ffff:x.x.x.x","worker":"xxxxx.xxx","difficulty":0.01,"error":"incorrect size of solution"}
And also there is no hashrate.
Unfortunately this is not the case with the problem. Not a single warning is presented in the log files.
Those ppl have good steady hasrate, everything is OK, just when it comes to finding block it got rejected from zcashd.
Just checked, the solution verification has stayed the same in optiminer since version 2.0. You are chasing a red herring here.
My best guess would be that there is NOMP and zerod validate the block difficulty.