# OverXelis Miner feedback

Do not include secrets, full payout addresses, private IPs, tokens, or unredacted configurations.

Version: 1.1.1
Edition: final
CPU model:
Linux distribution/version:
Python and GCC versions:
Build mode (normal non-PGO or experimental PGO):
Source-manifest and native-library SHA-256 (from build records):
Threads requested / active workers:
CPU order (automatic or explicit list):
Memory mode requested / confirmed (ordinary or hugetlb):
Warmup and measurement duration:
Local session average (including pauses):
Active-only average, when measured separately:
USER accepted / rejected / unresolved:
FEE accepted / rejected / unresolved:
Full USER -> FEE -> USER cycle observed?
Reconnects or pauses:
Expected behavior:
Observed behavior:
Steps to reproduce:
Short redacted log:

For comparisons, record both versions, build settings, memory, power, temperature and workload. Include all runs, exclusions, accepted difficulty-weighted work and complete fee-cycle accounting. A startup peak or failed run is not a benchmark.
