Benchmark tools (rev2) [kernel-images] by raiden-staging · Pull Request #97 · kernel/kernel-images

[benchmark] add benchmark endpoint and cdp runtime benchmark implementation

raiden-staging

committed

[benchmark] instrument wrapper to capture and export container startup timing and phases

raiden-staging

committed

[benchmark] benchmarks: use fixed-duration component benchmarks and report actual elapsed time

raiden-staging

committed

[benchmark] remove ffmpeg-downloader stage from chromium-headful dockerfile

raiden-staging

committed

[benchmark] remove screenshot benchmark support and related converters/types/api

raiden-staging

committed

[benchmark] integrate neko websocket trigger and client for webrtc benchmarks

raiden-staging

committed

[benchmark] adjust unikernel scripts and trim openapi benchmark docs

raiden-staging

committed

[benchmark] lower vcpu count to 2 for chromium-headful unikernel

raiden-staging

committed

[benchmark] benchmarks: switch from websocket neko trigger to reading continuous export; remove neko_client

raiden-staging

committed

[benchmark] add webrtc stats collector and stream benchmark payloads via websocket

raiden-staging

committed

[benchmark] wrap benchmark webrtc stats into payload field when sending over websocket

raiden-staging

committed

[benchmark] bench: make cdp scenarios session-aware and switch to generated commands

raiden-staging

committed

[benchmark] refactor cdp runtime benchmark to benchmark endpoints individually and simplify results

raiden-staging

committed

[benchmark] refactor cdp runtime benchmark and implement mixed-workload runner

raiden-staging

committed

[benchmark] refactor cdp runtime benchmark: create and attach a new page target per endpoint, inline target lifecycle and cleanup (detach/close)

raiden-staging

committed

[benchmark] cdp: remove flatten from Target.attachToTarget and add verbose diagnostics

raiden-staging

committed

[benchmark] simplify cdp benchmark by removing target creation/attach and session handling

raiden-staging

committed

[benchmark] fetch page websocket url from /json instead of /json/version

raiden-staging

committed

[benchmark] run mixed cdp workload with per-worker websocket connections

raiden-staging

committed

[benchmark] refactor cdp runtime benchmark: session-based workers, deterministic scenarios, improved stats and error sampling

raiden-staging

committed

[benchmark] timebox scenarios and record per-scenario duration

raiden-staging

committed

[benchmark] run scenarios in multiple rounds and increase default benchmark duration to 15s; compute per-scenario duration across rounds with 400ms min

raiden-staging

committed

[benchmark] simplify cdp runtime benchmark and session lifecycle

raiden-staging

committed

[benchmark] make cdp runtime benchmark sequential and run iterations per scenario

raiden-staging

committed

[benchmark] support per-scenario duration/iterations/timeouts and rework iteration loop

raiden-staging

committed

[benchmark] record attempt_count and duration_seconds in cdp scenario results and api

raiden-staging

committed

[benchmark] benchmarks/cdp: track cdp events, add scenario type and network fetch-burst; expose event counts/throughput in results

raiden-staging

committed

[benchmark] include scenario type and event metrics in cdp benchmark api output

raiden-staging

committed

[benchmark] sync with latest

raiden-staging

committed

[benchmark] update neko base version

raiden-staging

committed

[benchmark] mem tracking , phases deltas , system metrics

raiden-staging

committed

[benchmark] NaN guards

raiden-staging

committed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Benchmark tools (rev2) [kernel-images]#97

Benchmark tools (rev2) [kernel-images]#97
raiden-staging wants to merge 32 commits intokernel:mainfrom
raiden-staging:benchmark-wip-rev2

Commits on Nov 20, 2025

[benchmark] add benchmark endpoint and cdp runtime benchmark implementation

[benchmark] instrument wrapper to capture and export container startup timing and phases

[benchmark] benchmarks: use fixed-duration component benchmarks and report actual elapsed time

[benchmark] remove ffmpeg-downloader stage from chromium-headful dockerfile

[benchmark] remove screenshot benchmark support and related converters/types/api

[benchmark] integrate neko websocket trigger and client for webrtc benchmarks

[benchmark] adjust unikernel scripts and trim openapi benchmark docs

[benchmark] lower vcpu count to 2 for chromium-headful unikernel

[benchmark] benchmarks: switch from websocket neko trigger to reading continuous export; remove neko_client

[benchmark] add webrtc stats collector and stream benchmark payloads via websocket

[benchmark] wrap benchmark webrtc stats into payload field when sending over websocket

[benchmark] bench: make cdp scenarios session-aware and switch to generated commands

[benchmark] refactor cdp runtime benchmark to benchmark endpoints individually and simplify results

[benchmark] refactor cdp runtime benchmark and implement mixed-workload runner

[benchmark] refactor cdp runtime benchmark: create and attach a new page target per endpoint, inline target lifecycle and cleanup (detach/close)

[benchmark] cdp: remove flatten from Target.attachToTarget and add verbose diagnostics

[benchmark] simplify cdp benchmark by removing target creation/attach and session handling

[benchmark] fetch page websocket url from /json instead of /json/version

[benchmark] run mixed cdp workload with per-worker websocket connections

[benchmark] refactor cdp runtime benchmark: session-based workers, deterministic scenarios, improved stats and error sampling

[benchmark] timebox scenarios and record per-scenario duration

[benchmark] run scenarios in multiple rounds and increase default benchmark duration to 15s; compute per-scenario duration across rounds with 400ms min

[benchmark] simplify cdp runtime benchmark and session lifecycle

[benchmark] make cdp runtime benchmark sequential and run iterations per scenario

[benchmark] support per-scenario duration/iterations/timeouts and rework iteration loop

[benchmark] record attempt_count and duration_seconds in cdp scenario results and api

[benchmark] benchmarks/cdp: track cdp events, add scenario type and network fetch-burst; expose event counts/throughput in results

[benchmark] include scenario type and event metrics in cdp benchmark api output

[benchmark] sync with latest

[benchmark] update neko base version

[benchmark] mem tracking , phases deltas , system metrics

[benchmark] NaN guards