New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

WIP: provide timeseries marking editor for backtesting #1569

wants to merge 3 commits into
base: develop


None yet
2 participants

Dieterbe commented Mar 6, 2015

the idea is to build a UX to easily annotate historical data as ok/warn/crit.
this means that software that does alerting, fault detection and machine learning can be validated against this historical data (backtesting).
we can even automatically score alerting logic by computing how well its output matches the annotated historical information.
here's a 45second video that quickly shows the UX idea in action:

works :

  • selecting frames, marking as warn/ok/... and retaining the state throughout zoom operations. i think i've been able to come up with a fairly elegant way to differentiate against zoom, and mark as ok/warn/crit (for the latter 3, the mouseup should be in the corresponding bar on top)

i'm a newcomer to the grafana codebase, and to UX, i know the current code is not great. i welcome feedback to make this better. in specific i'm currently trying to figure out:

next up:

  • deleting or adjusting existing annotations (no idea yet how to do this)
  • loading and saving annotations to/from ES (slowly reverse engineering codebase)
  • currently the markings are saved in the dashboard definition, this is not good. they should only go in an external service such as ES.
  • coming up with a way to build out this UX on a per-alerting rule basis (with 1 or more useful series), and leave other panels alone ;)

Dieterbe added some commits Mar 9, 2015

show active timerange markings in tooltip
a bit expensive perhaps, but should become useful to delete timeranges

@torkelo torkelo closed this Apr 21, 2015

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment