-
Notifications
You must be signed in to change notification settings - Fork 24.9k
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
Deprecation warnings without an obvious cause or mitigation #81589
Comments
Pinging @elastic/es-security (Team:Security) |
Pinging @elastic/es-core-infra (Team:Core/Infra) |
I had a brief look at the case of And I can confirm it does issue deprecation warning if the user has access to system indices. The reason is that GetAliasRequest sees the above request to be equivalent to Lines 383 to 385 in 8eb23b6
It then issues warnings based on all of the authorized aliases without taking in account of the actual index pattern ( Line 198 in 8eb23b6
The straightforward fix seems to be restricting the checking to only aliases that have actually been resolved for the requested index patterns. In fact, the check for system indices (not aliases) access does filter out for the requested indices only. |
This is from the use of internal monitoring, which is deprecated but still in use pending a migration to metricbeat monitoring. I will follow up to see if we should hide that deprecation warning on platforms still impacted. |
OK... it look like |
Part of this is that good ol' thing Security does of replacing the index patterns in the request with a list of authorized indices, so we can't do anything based on the actual index pattern because we don't have it. Anyway, what this check is doing is checking for requests that try to access aliases that have names reserved for system indices/aliases. I'm not sure why we were doing that, as we don't try to do that anywhere else - we only warn about accessing things that actually exist, rather than things that would be problematic to access if they did exist. Anyway, I think the solution here is just to remove the warning entirely and I'll have a PR up to do so shortly. |
The check removed by this commit checks specifically for requests for aliases that *would* be system aliases, if they existed. I'm not sure why we were doing this, as we don't try to do this anywhere else. The only test that this seems to make fail is the test explicitly checking that behavior, which I don't think is really what we want. So I'm just removing it. Relates #81589
The check removed by this commit checks specifically for requests for aliases that *would* be system aliases, if they existed. I'm not sure why we were doing this, as we don't try to do this anywhere else. The only test that this seems to make fail is the test explicitly checking that behavior, which I don't think is really what we want. So I'm just removing it. Relates elastic#81589
) The check removed by this commit checks specifically for requests for aliases that *would* be system aliases, if they existed. I'm not sure why we were doing this, as we don't try to do this anywhere else. The only test that this seems to make fail is the test explicitly checking that behavior, which I don't think is really what we want. So I'm just removing it. Relates #81589
The check removed by this commit checks specifically for requests for aliases that *would* be system aliases, if they existed. I'm not sure why we were doing this, as we don't try to do this anywhere else. The only test that this seems to make fail is the test explicitly checking that behavior, which I don't think is really what we want. So I'm just removing it. Relates elastic#81589
) The check removed by this commit checks specifically for requests for aliases that *would* be system aliases, if they existed. I'm not sure why we were doing this, as we don't try to do this anywhere else. The only test that this seems to make fail is the test explicitly checking that behavior, which I don't think is really what we want. So I'm just removing it. Relates #81589
As noted above, some of these are expected, and others (the |
After reviewing the deprecation logs from Kibana CI elastic/kibana#120043 I came across several requests which triggered deprecations around system indices even though the request doesn't explicitly access system indices. Are these false positives? Could we provide a message that helps users resolve these?
The text was updated successfully, but these errors were encountered: