Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Global settings #3933

Closed
4 tasks done
Tracked by #3428
zhang2014 opened this issue Jan 21, 2022 · 3 comments · Fixed by #6579
Closed
4 tasks done
Tracked by #3428

Global settings #3933

zhang2014 opened this issue Jan 21, 2022 · 3 comments · Fixed by #6579
Assignees
Labels
A-executor Area: processor and shuffle C-feature Category: feature

Comments

@zhang2014
Copy link
Member

zhang2014 commented Jan 21, 2022

SET GLOBAL max_block_size = 30;
-- We should store the settings in meta service and get this for cluster other nodes of the tenant.

Tasks:

  • Refine the session: change mutable status to session context
  • Add settings set in meta
  • Write/Read global settings for settings
  • SET GLOBAL xx=yy statement
@BohuTANG
Copy link
Member

BohuTANG commented Feb 3, 2022

From the experience of snow, it seems that the settings should bind with tenant not the cluster, from the user side, he shouldn't know what's the cluster is.
So, what is the purpose related to the cluster setting here?

@zhang2014
Copy link
Member Author

zhang2014 commented Feb 3, 2022

From the experience of snow, it seems that the settings should bind with tenant not the cluster, from the user side, he shouldn't know what's the cluster is.
So, what is the purpose related to the cluster setting here?

This can be useful when opening the features of different warehouses. For example, on user side, the test cluster (read-only) and the stable cluster may share the same data, but they have different settings. But it seems that global is more suitable for tenants.

Possible solutions(multiple level settings):

SET GLOBAL enable_some_experimental_feature = false; -- tenant
SET WAREHOUSE enable_some_experimental_feature = true; -- cluster(higher priority)

I misunderstood. you are right. in this case, we may use the shared data feature(different tenant). SET GLOBAL should bind with tenant level.

@BohuTANG BohuTANG changed the title Cluster level global settings Global settings Feb 3, 2022
@BohuTANG BohuTANG added A-executor Area: processor and shuffle and removed A-query Area: databend query labels Mar 29, 2022
@BohuTANG
Copy link
Member

BohuTANG commented Jun 27, 2022

I think we can restart the task now.
cc @zhang2014 @sundy-li

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
A-executor Area: processor and shuffle C-feature Category: feature
Projects
3 participants