LATENCY command and related subcommands

yes

trying that now

admin@dragonfly-testing:~$ memtier_benchmark -h dragonfly.***.*** --ratio 1:0  -n  10000
Writing results to stdout
[RUN #1] Preparing benchmark client...
[RUN #1] Launching threads now...
[RUN #1 100%,  10 secs]  0 threads:     2000000 ops,  204977 (avg:  199985) ops/sec, 15.42MB/sec (avg: 15.05MB/sec),  0.97 (avg:  1.00) msec latency

4         Threads
50        Connections per thread
10000     Requests per client


ALL STATS
============================================================================================================================
Type         Ops/sec     Hits/sec   Misses/sec    Avg. Latency     p50 Latency     p99 Latency   p99.9 Latency       KB/sec
----------------------------------------------------------------------------------------------------------------------------
Sets       193587.77          ---          ---         0.99960         0.97500         2.07900         4.95900     14913.65
Gets            0.00         0.00         0.00             ---             ---             ---             ---         0.00
Waits           0.00          ---          ---             ---             ---             ---             ---          ---
Totals     193587.77         0.00         0.00         0.99960         0.97500         2.07900         4.95900     14913.65

Awaiting metrics in grafana and will update shortly

initial set latency (milliseconds unit) pointing to the same range

which I don’t really trust, I think memtier benchmark is a better source of truth at the moment

yes your memtier numbers looks correct

if the latency is 200ms the numbers would be quite different

definitely

I will ask other guys what can be the reason with grafana, if find the answer let you know

sounds great, thanks for all the help

and version just in case:

admin@dragonfly:~$ dragonfly --version
dragonfly v1.12.1-053a33d24df939d1c35f3b6f233a0c57ac185184
build time: 2023-11-06 11:40:00

@reliable-impala Could you also check what latency shows by dragonfly latency command

I was told that grafana uses server latency and should be less than memtier because we don’t have wastes on networking

yes, i’ll do further testing and see if i can get you an answer by monday

maybe somehow grafana shows latency in the incorrect measure

is there a specific latency command that you want me to run?

just simple latency command that you have mentioned

will do

I think instead of 250ms it should be 2.5 ms or 250us

if I right understand how grafana works)