-
Notifications
You must be signed in to change notification settings - Fork 6
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
Danielh/eng 833 coalesce multiple opsani dev remedies to #411
Danielh/eng 833 coalesce multiple opsani dev remedies to #411
Conversation
ENG-833 Coalesce multiple Opsani Dev remedies to reduce user app churn
The current approach used by the opsani_dev connector causes the user applicatoin to restart 3(!) times, even though once is enough (if all 3 remedies were applied at the same time). I consider this a bug as it causes problems for customers (and our support team), especially when the application has a high number of replicas (which means an interesting application) and/or starts slowly. If the opsani_dev-based onboarding will continue to be used, this will need to be resolved. I acknowledge that fixing this may require design changes/modifying the approach to checks-and-remedies. See some discussion at https://linear.app/opsani/issue/SVX-274/see-about-coalescing-multiple-hints as a starting point. |
Basically there isn't a need for the full readiness check after every remedy, as I see it it's primarily important just to have a required halting prior to the tuning pod check to ensure everything was succesfully ingested before its creation |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
from a cursory glance a couple of items:
- I dont see any automated unit/integration tests added that cover both types of the configuration
- You added a CLI flag for the checks which is good but I thought there was some kind of check configuration section for the servo.yaml. Can we get it added there as well as it may be required for environment variable configuration
@linkous8 I haven't found a check configuration outside of the CLI flag. Also I don't see any tests that cover the flags of the relevant CLI section, will see about adding that. Do you think an integration test running the two variants of checking on say, fiber-http would be good? |
Lets go ahead and add a new ChecksConfiguration model for that, say, around here: Line 467 in d027184 Long term, I'm looking to streamline the CLI entrypoint to make the flow more understandable and will be preferring file/env var based configuration over CLI flags as the latter is more difficult to keep track of and alter programmatically
See if you can adapt this test to test both variants of the configuration: servox/tests/connectors/opsani_dev_test.py Line 1176 in d027184
|
…eck halting integration tests, pull string functions out to utilities
…load_kubeconfig on k8s check
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
On the right track 🚂
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
A few minor nitpicks aside, LGTM! 🚀
No description provided.