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

Throttling is too aggressive #1572

Closed
jacobthill opened this issue Aug 31, 2022 · 12 comments
Closed

Throttling is too aggressive #1572

jacobthill opened this issue Aug 31, 2022 · 12 comments
Assignees

Comments

@jacobthill
Copy link
Contributor

I believe throttling was implemented as part of our efforts to improve site performance. I don't know if the settings are different in different environments. As the DLME data manager, I use the development environment for quality checking data transformations. Normal activity for me triggers our throttling which results in a screen that says "Throttled".

This impedes data aggregation work. Can we revisit the original intent behind this and come up with a better solution? Is this implemented in all environments or just dev? Does this have anything to do with bots? If not, humans don't typically use this site except Stanford devs and the DLME data manager.

@corylown
Copy link
Contributor

I think it's likely we can close this issue since we're moving the web application on premise, but leaving it open for evaluation during the spring 2023 work cycle.

@jcoyne
Copy link
Contributor

jcoyne commented Jan 25, 2023

Might be rack-attack

@thatbudakguy
Copy link
Member

deferring this until on-prem migration is mostly complete; then we can test again to see if the situation has changed

@jacobthill
Copy link
Contributor Author

The issue persists after the migration on prem.

@thatbudakguy
Copy link
Member

moving to "ready"; this needs more analysis to determine the root cause.

@thatbudakguy
Copy link
Member

@jacobthill can you describe activity that reliably triggers the "throttled" message for you? right now, it looks like the only scenarios we throttle are:

  • you try to login more than 20 times in one hour
  • you make more than 15 searches in a 15-minute period, where:
    • there is no search term, and
    • you are requesting a page of results deeper than page 10

are these things that you do frequently?

@jacobthill
Copy link
Contributor Author

Its always filtering via facets. When I load a data set like AUC, for example, I need to select each collection then check that the Type facets are correct. So I'm selecting "Data provider" > "Collection id" > "Type", then expanding "Type" for all collections.

@thatbudakguy
Copy link
Member

got it, thanks for explaining. and do you only do this in stage? or are there scenarios where you would do it in prod as well?

@jacobthill
Copy link
Contributor Author

Only in stage.

@thatbudakguy
Copy link
Member

confirmed disabled in stage; re-enabled in prod. closing this.

@jacobthill
Copy link
Contributor Author

I am reopening this issue because the user behavior is now changed. Some curatorial work that was previously done in stage is now done in prod–the same kind of faceting described above "So I'm selecting "Data provider" > "Collection id" > "Type", then expanding "Type" for all collections." We need to fill out the collections registry form (a spreadsheet with collection level metadata) and we will have multiple users working on this at the same time. Currently, we can only work for 5-10 minutes before we get throttled. We need to remove/adjust throttling in prod.

@jacobthill
Copy link
Contributor Author

Closing based on #1864

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
No open projects
Status: Todo
Development

No branches or pull requests

4 participants