Skip to content

testing.md

Latest commit

 

History

History
94 lines (62 loc) · 3.24 KB

testing.md

File metadata and controls

94 lines (62 loc) · 3.24 KB

Testing

Default Suite

The default test suite uses in-process smart HTTP servers and does not require a local listener:

env GOCACHE=/tmp/go-build go test ./...

git-http-backend End-To-End Tests

Optional end-to-end write test against the system git-http-backend:

env GOCACHE=/tmp/go-build GITSYNC_E2E_GIT_HTTP_BACKEND=1 go test ./internal/syncer -run TestRun_GitHTTPBackendSync -v

That path exercises real smart HTTP fetch and push with a local bare source repo and a local bare target repo.

Dedicated batched bootstrap coverage:

env GOCACHE=/tmp/go-build GITSYNC_E2E_GIT_HTTP_BACKEND=1 go test ./internal/syncer -run TestBootstrap_GitHTTPBackendBatchedBranch -v

Batch-planning sensitivity coverage:

env GOCACHE=/tmp/go-build GITSYNC_E2E_GIT_HTTP_BACKEND=1 go test ./internal/syncer -run TestBootstrap_GitHTTPBackendBatchedPlanningTracksBatchLimit -v

That test uses a real git-http-backend source/target pair and checks that a smaller --target-max-pack-bytes planning limit produces at least as many planned checkpoints as a larger one, while still planning to the branch tip.

Live Linux Smokes

Optional live Linux bootstrap smoke:

env GOCACHE=/tmp/go-build GITSYNC_E2E_LIVE_LINUX=1 go test ./internal/syncer -run TestBootstrap_LiveLinuxSource -v

Batched variant:

env GOCACHE=/tmp/go-build GITSYNC_E2E_LIVE_LINUX=1 go test ./internal/syncer -run TestBootstrap_LiveLinuxSourceBatched -v

These are useful for large-source relay and memory checks while keeping the target local and disposable.

Benchmarking

git-sync-bench runs repeatable empty-target benchmarks against a source repository. It creates a fresh bare target for each run and reports wall-clock time plus internal syncer measurement data.

Build it with:

go build -o /tmp/git-sync-bench ./cmd/git-sync-bench

Example against a local mirror:

/tmp/git-sync-bench \
  --scenario bootstrap \
  --source-url /tmp/git-sync-bench/kubernetes.git \
  --repeat 3 \
  --target-max-pack-bytes 104857600 \
  --stats \
  --json

The JSON report includes per-run results, aggregate min/avg/max timings, batch counts for batched runs, heap peaks, and relay modes seen across successful runs. If --source-url is a filesystem path, the tool converts it to file://... automatically.

Use --keep-targets to retain generated bare targets under --work-dir for inspection. For large real-repo runs, prefer a local mirror instead of benchmarking directly against a hosted remote.

mise Tasks

  • mise run test — default suite
  • mise run test:ci — default suite with race detection
  • mise run test:git-http-backend — git-http-backend end-to-end
  • mise run test:linux-smoke — live Linux bootstrap smoke
  • mise run test:linux-smoke:batched — live Linux batched bootstrap smoke

test:linux-smoke and test:linux-smoke:batched use a disposable local bare Git target served through git-http-backend. They do not require any external service beyond the public source remote.

TLS Overrides

For local or self-signed targets:

  • --source-insecure-skip-tls-verify
  • --target-insecure-skip-tls-verify
  • GITSYNC_SOURCE_INSECURE_SKIP_TLS_VERIFY=true
  • GITSYNC_TARGET_INSECURE_SKIP_TLS_VERIFY=true