search

als-computing / splash-deploy

Other
1 stars 3 forks source link

Index databroker start document in elastic #8

Open dylanmcreynolds opened 4 years ago

dylanmcreynolds commented 4 years ago

Currently, the system indexes an "experiment" which was a nice idea but really only good for a demo. While the concept of "experiemnt" may live on, what that model looks like is a little unclear.

One thing is clear, though. Splash is going be consuming bluesky runs as the principal place to store "run" data...loosely defined as the stuff that gets store for each trial in an experiment. https://blueskyproject.io/databroker/v2/user/index.html

There is a well-defined schema for a set of document types in the bluesky event model (https://github.com/bluesky/event-model/tree/master/event_model/schemas). We will get a lot of useful functionality out of these documents if we index the run_start documents in elastic as they come into mongo.

The end goal is to make it easy to search for a run based on several fields of particular interest (but not exclusively): time user bealline_id sample detectors

So, this task is to set up create the stuff in splash-deploy that us perform smart searches (fuzzy, suggestions, etc) on those fields in start documents.

J-avery32 commented 4 years ago

Finished the mapping configuration. What's left is to make the front end compatible with this format now.

dylanmcreynolds commented 4 years ago

Do you have a pull request? Note that if you create one, you can reference this issue in the comments and github will auto-close this issue when the PR is merged.