keryx-miner-supr v0.6.5.6
Adds automatic pool failover — opt-in and, with no backup configured, behaviour is byte-identical to previous versions.
--backup-pool (repeatable → priority list)
-sis the primary; each--backup-pool stratum+tcp://host:portis a lower-priority backup, tried in the order given.- If the current pool fails hard — no job for ~90 s (
KERYX_FAILOVER_AFTER_SECS) or the connection drops — the miner fails over to the next pool in the list. - While off the primary, it keeps probing higher-priority pools in the background and fails back to the highest one that recovers (grace
KERYX_FAILBACK_GRACE_SECS=60 s, probeKERYX_FAILBACK_PROBE_SECS=30 s), with anti-flap backoff. Uses the same wallet/password as-s.
Example: -s stratum+tcp://poolA:4401 --backup-pool stratum+tcp://poolB:4401 --backup-pool stratum+tcp://poolC:4401
Live-validated (failover + failback, 0 rejects). Nothing in the mining hot path or the no-backup path was changed.
Build lines
modern (CUDA 12.9, R575+), legacy (CUDA 12.4, R550+), pascal (sm_60; --cpu-inference --light). Each in hiveos / smos / mmpos / linux.