refactor: wrappers for cache - #489
Conversation
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## main #489 +/- ##
==========================================
- Coverage 71.33% 71.21% -0.12%
==========================================
Files 233 234 +1
Lines 18999 19028 +29
==========================================
- Hits 13552 13551 -1
- Misses 4419 4438 +19
- Partials 1028 1039 +11 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
|
@seqbenchbot start bulk main --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start bulk 0-scoped-cache --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start mixed main --duration=15m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start mixed 0-scoped-cache --duration=15m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start mixed 0-scoped-cache --duration=15m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start mixed main --duration=15m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot compare f0f3659f 785f5d30 |
|
Nice, @dkharms Comparison of baseline Show comparison
See both runs overlaid (baseline aligned to comparison) on the Grafana dashboard. Have a great time! |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
@seqbenchbot start search-regular main 'message:*' --duration 3m |
|
Nice, @cheb0 Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @cheb0 Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start search-regular 0-scoped-cache 'message:*' --duration 3m |
|
Nice, @cheb0 Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @cheb0 Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start bulk 0-scoped-cache --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start bulk main --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start bulk 0-scoped-cache --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
|
@seqbenchbot start bulk 0-scoped-cache --duration=10m |
|
Nice, @dkharms Started benchmark Show details
I'll post the statistics here once it finishes. Here is a list of helpful links: Have a great time! |
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
|
Nice, @dkharms Benchmark Show summary
Compare it against another run with |
9262aea to
3cf7a7e
Compare
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
3cf7a7e to
76b8df4
Compare
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
76b8df4 to
7738f33
Compare
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
7738f33 to
39a1693
Compare
🔴 Performance DegradationSome benchmarks have degraded compared to the previous run. Show table
|
Session and Scan wrappers for cache
Description
In this PR I've introduced interface
cache.Wrapperintocachepackage.This interface allows us to solve several problems by providing the most reasonable implementations at the call-sites. The hottest problems are:
lidscache layer) because we receive objects one-at-a-time (from the one block) and therefore generate a lot of cache hits;It was a little bit tricky to collect sound measurements because compaction is using
cache.Scanwrapper and the overall RAM usage dropped and therefore GC was more sensitive to RAM usage spikes (example, GC-rate) that's why I've setGOGC=500andGOMEMLIMIT=4GiB.So I guess the most important metrics are:
0-scoped-cachebranch because compaction does not evict cached blocks and search latencies are better across all percentiles;mean (ms)stddev (ms)p(50) (ms)p(95) (ms)p(99) (ms)iterationstotal25.8425.81-0.15%5.325.20-2.37%25250.00%3836-5.26%5953-10.17%3232932328-0.00%000.00%43.3841.80-3.65%5.015.16+2.86%43.2841.55-4.00%56.5954.35-3.96%75.8770.06-7.66%478547850.00%000.00%If you have used LLM/AI assistance please provide model name and full prompt: