Not right now. @User is adding support for text search

e.g.

companyId, Name,  Description
1234, foo, "this is an awesome company"

you can model them as

companyId, Name,  Description, Description_tags
1234, foo, "this is an awesome company", ["this", "is", "awesome", "company"]

then you can do something like description_tags IN ('awesome') and description_tags IN ("company")

Cool. I'll try this.

good morning! What is the correct way to swap offline segments (lets say when we want to republish them as we missed some data or if we added new column and want to backfill) ?

If old and new segment names match, simply pushing will overwrite the old segments (aka refresh).

It’s a bit tricky if old segment names and number don’t match, and you want to overwrite all data without taking downtime

@User ^^

if I overwrite the file, do I need to call API again to register it?

or I can just write to a DFS?

@User?

You need to push the segments to the controller again

Unless if you are just adding a new column and you want to provide a default value for existing segments, in that case you just update schema and call the back fill api to populate default value for new column @Alex

got it, thanks.

@User just wondering regarding this PR (https://github.com/apache/incubator-pinot/pull/4952) if you're ok with going ahead with the tenant API - tagging support ?

i will chat with snlee

ok lemme know - I can move those endpoints to

/tenant

instead of

/instance

will do

cc: @User @User @User this is awesome!

<!here> Thanks for being part of the Pinot community. As you all know, Pinot is part of Apache Incubator and one key requirement to graduate to a top level project is to have a strong community. Apache measures this based on the activity in apache mailing lists (users@pinot.apache.org) (unfortunately, slack messages are not considered part of this). Request every one here to subscribe to the Apache mailing list. This is as simple as sending an email to users-subscribe@pinot.apache.org. We will still have our conversations on slack and we can use slack plugins to redirect messages to Apache mailing list.

Slack activity graph.

For folks not aware of Pinot's twitter handle: https://twitter.com/ApachePinot

It will be good to tweet the announcements/blog posts etc

<!here>, text search PR

Question - we just updated pinot to head and have realtime tables. Since the jar locations changed (i.e. KafkaCopnsumerFactory) for the realtime tables I resubmitted the curl statement to add a realtime table, is that all that needs to be done? I looked in "externalView" and do not see any errors.

yes, thats all thats needed. we should probably have a mapping from old class to new class to make the migration simpler. wdyt?

Makes sense, would be easier for users to upgrade. Or atleast fail the new segments (i.e. new online segments in ERROR state in externalView) and have a descriptive error message so users know what to do?