add print frequency

[wanghenshui]

log message is to much when encountered an connect error and I add this option to control print frequency

[wanghenshui]

ping @yossigo

[yossigo]

@wanghenshui thank you! Just be sure I understand correctly, so you basically want the ability to bump up the print rate to overcome lots of messages produced by benchmark_*_log() prints on connection errors, etc.?

Did you consider an option to suppress those messages altogether, possibly narrowing it down to an additional errors counter?

[wanghenshui]

@wanghenshui thank you! Just be sure I understand correctly, so you basically want the ability to bump up the print rate to overcome lots of messages produced by benchmark_*_log() prints on connection errors, etc.?

Did you consider an option to suppress those messages altogether, possibly narrowing it down to an additional errors counter?

we use memtier_benchmark as a stress tool, stress our redis-like server every day, print a log per second is also too much for analyze.

narrowing it down to an additional errors counter is also a good idea, maybe i implement it in next PR