Replies: 1 comment 1 reply
-
|
You're more than welcome to open a new SIP (I'll convert this to a Discussion rather than an Issue in the meantime). I think that'll be cleaner than reopening the old one, especially if you want to own putting it through the SIP process. That said, I think you'll likely run into similar problems mentione on the other SIP... namely:
Database agnosticism and opt-in deployment do sound like important facets though. There are any other number of things that come through the logging service that people may or may not want in their monitoring/activity dashboards. I guess this all points to the question of if/how it can be configurable/customizable. I have nothing against any of this, just don't know how to make it perfect/universal. I look forward to the SIP if you want to go down that road :) |
Beta Was this translation helpful? Give feedback.
Uh oh!
There was an error while loading. Please reload this page.
-
Hi @rusackas @TechAuditBI,
SIP-122 (#27620) proposed an integrated monitoring and user activity dashboard, but it was closed due to inactivity and is now locked for comments so I'm opening this to continue the conversation.
I've searched the repository, closed PRs, and the SIP archive, and I couldn't find any implemented work or follow-up related to this proposal as far as I can tell, this gap still exists and community discussions (e.g. #33538) keep pointing back to that issue.
This is a problem I already solve today at my organization by manually extracting data from the metadata database, so I have practical experience with it and I'd like to move this proposal forward.
My plan is to open a new SIP referencing SIP-122, with a complete specification that addresses the points raised in the original discussion:
Metadata-database-agnostic approach (no PostgreSQL-specific materialized views supporting both PostgreSQL and MySQL metastores)
Scope limited to user activity and content usage (explicitly excluding infra metrics like CPU/memory)
Admin-only access by default
Optional, opt-in deployment (e.g. superset load_monitoring), separate from example dashboards
A clear migration plan
Before I write it up: would the maintainers prefer a new SIP, or reopening #27620?
Beta Was this translation helpful? Give feedback.
All reactions