How can I cleanup events?

[joel-apollo]

Issue Summary

We've been running Redash for several years now and our events table has accumulated ~94 GB of data across 5 years. I tried looking in the docs, but there's no mention of how this can be managed or cleaned up. I definitely don't feel like I need to keep 5 years of access events. At the same time, it's valuable to know when someone created or last edited a query.

Steps to Reproduce

I expect the admin manual to describe steps to cleanup old events data

Technical details:

• Redash Version: Now on
• Browser/OS:
• How did you install Redash: Docker on ECS & RDS

[arikfr]

For the SaaS Redash what I was doing is periodic pruning of the table using something like the following SQL script:

BEGIN;

-- Create new table with proper structure (CREATE TABLE ... AS doesn't copy constraints/defaults)
CREATE TABLE events_new (LIKE events INCLUDING ALL);
INSERT INTO events_new SELECT * FROM events WHERE created_at > current_date - interval '90 days';

-- Swap
ALTER TABLE events RENAME TO events_old;
ALTER TABLE events_new RENAME TO events;

-- Fix the sequence ownership (outside transaction is fine)
ALTER SEQUENCE events_id_seq OWNED BY events.id;
ALTER TABLE events ALTER COLUMN id SET DEFAULT nextval('events_id_seq');

-- Grants
GRANT ALL ON events TO redash;
GRANT SELECT ON events TO redash_reader; -- you might not have this user in your deployment, so can be skipped

COMMIT;

-- Once verified, drop the old table
-- DROP TABLE events_old;

[joel-apollo]

Thanks, that's helpful!

[arikfr]

Please share how it went for you and then we can create a doc page about this based on your experience.

It might be worth to create a cleanup job like we have for query results.