[MDS] datasource selector added #29
Conversation
Signed-off-by: Kawika Avilla <[email protected]>
Signed-off-by: Kawika Avilla <[email protected]>
Signed-off-by: Kawika Avilla <[email protected]>
Signed-off-by: Kawika Avilla <[email protected]>
| useEffect(() => { | ||
| const subscriptions = connectionsService.getSelectedConnection$().subscribe((connection) => { | ||
| connectionsService.setSelectedConnection$(connection); | ||
| }); | ||
| return () => { | ||
| subscriptions.unsubscribe(); | ||
| }; | ||
| }, [connectionsService]); |
There was a problem hiding this comment.
why is this getting connection and setting the same connection?
There was a problem hiding this comment.
good catch. whenever Discover page just incase the client switched off page i can still set it in the service. but tbh it might just be left over. will look again post merge
There was a problem hiding this comment.
didn't fully understand.. but i just tested this PR and this section is giving me Uncaught RangeError: Maximum call stack size exceeded
| const resp = await context.core.savedObjects.client.find<DataSourceAttributes>({ | ||
| type: 'data-source', | ||
| fields, | ||
| perPage: 10000, |
There was a problem hiding this comment.
it could. and probably should once we start adding more to this wrapper
| title: string; | ||
| endpoint?: string; | ||
| installedPlugins?: string[]; | ||
| auth?: any; |
There was a problem hiding this comment.
There was a problem hiding this comment.
yeah it should. the original approach i wanted to go with is not requiring any imports from the data source plugin in case it didn't get loaded in browser (data_source.enabled: false) because i dont think they have any lazy loading over there.
i think my goal here is even if we want to take the work for the flint datasources stuff can live in this repo but extend what they have in data sources.
where we can properly lazy loaded and make it a lot easier to use and then eventually unify it with the ds mgmt plugin
There was a problem hiding this comment.
i see. i thought it wouldn't matter for types and import/import type {xxx would be removed when transpiled to js
Co-authored-by: Joshua Li <[email protected]>
Co-authored-by: Joshua Li <[email protected]>
Co-authored-by: Joshua Li <[email protected]>
Signed-off-by: Kawika Avilla <[email protected]>
Added data source selector that pulls from saved objects and also added a connections service and passing it with the dataframe.
Related to:
opensearch-project/OpenSearch-Dashboards#7157
With DQL (and Lucene):
With PPL (and SQL):
