Pavel Borisov

Non-s3 mode. Database 6Gb (I limited it from previous mesaurements, otherwise it's slow to fill). 10000 partitions Shutdown (5 min) takes 2 long checkpoints (checkpointer process @ ~25% CPU): First...

S3 mode. Database 6Gb. 10000 partitions Shutdown 2.5-5min (probably depend on current cluster state): checkpointer @50% CPU + many s3 workers @ ~20% CPU each   Startup: 1 min...

On the same s3 db as previous measurements: **1st Startup time:** 53s counting partitions @ 100% CPU (part1) 1m:49s cleanup old file @ 2% CPU (part2) **1st Shutdown time** 35...

Wallclock (gdb) flamegraph same as https://github.com/orioledb/orioledb/issues/328#issuecomment-1870347015 taken at be17f3abdff23f2e (last commit without segfault #333) start1 (counting partitions @100%)  start2 (cleanup @2%)  stop (checkpointer @1%)  stop (s3 worker...

https://github.com/orioledb/orioledb/commit/234115c782dd86b910a87cfd782bbec18d2f9361 experiences segfault https://github.com/orioledb/orioledb/issues/333 so the cluster could not be filled with data in a reproducible way. I'd suggest fixing #333 then recheck

at [02bc078](https://github.com/orioledb/orioledb/commit/02bc078a30f71940500be7eb63ff86badb17a3fa) startup: 60s (counting partitions) shutdown 10s

I noticed no difference in CI build to https://github.com/orioledb/orioledb/commit/bd7bb416864884a7ba5f9d14c8cc713e76e89e30