Skip to content

v0.1.6 — Span List Query Speedup + Endpoint Self-Latency Breakdown + databuff-proxy Dual-Write Migration

Choose a tag to compare

View all tags
@databufflabs databufflabs released this 03 Aug 01:32
· 78 commits to master since this release
v0.1.6
05ccbeb

Highlights

  • databuff-proxy dual-write migration — v0.1.6 refines the SkyWalking→DataBuff dual-write migration path and stability validation. Zero business-side changes (no agent swap, no config edits, no pod restarts); deploy databuff-proxy on the OAP host to mirror the same trace stream into both OAP and DataBuff Ingest. proxy is a separate repo: https://github.com/databufflabs/databuff-proxy
  • Endpoint self-latency breakdown — Endpoint analysis adds an "interface self duration" dimension that splits self-latency from downstream calls (DB / Redis / MQ / HTTP / RPC), so you can tell at a glance whether slowness is in the interface logic or a downstream dependency.
  • Span list query speedup — Span lists in endpoint analysis / trace search now use exact URL matching, drop the separate total count, and prune by a 30-minute startTime partition window for faster list responses.
  • Mutation (YoY/WoW) alarm direction fixed — Mutation alarms now respect comparePeriod and fluctuate direction (valUp / valDown / yoyUp / yoyDown), no longer misjudging direction.
  • Alarm tag values & eval SQL — listTagValues skips unknown metric tags; eval SQL covers filter operators.
  • AI stability — Brain multi-step handoff prompts improved, protocol coaching stripped out; session list/count is DB-first with fail-closed to memory on Doris errors; model failures surface to the user; brain dispatch bug fixed; brain routing prompt refreshed.
  • AI stability testing — All 5 AI integration suites run in parallel (chat / formats / memory / brain + new modelfail); new expert-interaction contract tests added; fixes for memory isolation, model-failure surfacing, and brain dispatch.
  • Workspace file preview — AI chat workspace online preview extended to all non-binary files (no longer limited to txt/log/csv/json).
  • Build & deploy — compose health wait timeout raised to 300s; new ai-apm-web-base image lets web release skip apt reinstalls.

Quick Start

# Platform (latest = 0.1.6)
curl -fsSL https://databuff.ai/databuff/ai-apm-install.sh | bash

# Upgrade existing Docker install (keeps data/)
curl -fsSL https://databuff.ai/databuff/ai-apm-update.sh | bash

Open http://YOUR_HOST:27403 · login admin / Databuff@123 · configure LLM API key for AI features.

Upgrade Notes

No schema migration in this release — upgrade swaps images and restarts. Kubernetes does not support in-place upgrade this release; uninstall then reinstall. See deploy/docker/UPGRADE.md.

Images

Component Image
Web Platform databuffhub/ai-apm-web:0.1.6
Ingest databuffhub/ai-apm-ingest:0.1.6
Demo databuffhub/ai-apm-demo:0.1.6

Links

Full changelog

See CHANGELOG.md.

Assets 2
Loading