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

Ambassador config for rolling updates #294

Closed
ukclivecox opened this issue Nov 12, 2018 · 10 comments
Closed

Ambassador config for rolling updates #294

ukclivecox opened this issue Nov 12, 2018 · 10 comments

Comments

@ukclivecox
Copy link
Contributor

When doing a rolling update of a model the kubernetes endpoint might change and requests passing through ambassador should handle the new ip-address without service-504 errors. One suggestion from Ambassador Slack channel is to use an Ambassador config like:

     timeout_ms: 0
     envoy_override:
       retry_policy:
         retry_on: connect-failure
         num_retries: 4

This needs to be tested and if works we should decide if its added by default.

@stefansedich
Copy link

@cliveseldon this is something I just ran into myself, is there an easy way to set this for my testing or is it something that needs to be done within seldon itself?

@stefansedich
Copy link

Ignore me @cliveseldon I see where it is now!

@ukclivecox
Copy link
Contributor Author

Yes, you would need to add to update the Ambassador annotation in the created Service. We'd love to hear your feedback if it helps you so we can increase priority to do this.

@stefansedich
Copy link

stefansedich commented Feb 17, 2019 via email

@ukclivecox
Copy link
Contributor Author

@stefansedich Can you confirm your use of these Ambassador settings solved the issue for you?

@stefansedich
Copy link

@cliveseldon I have not had a chance to test it out yet, will let you know how it goes when I get some time to do it.

@ukclivecox
Copy link
Contributor Author

It looks like for Ambassador 0.5 release envoy_override has been removed for now so this functionality is only available for 0.4: See emissary-ingress/emissary#1127

@ukclivecox
Copy link
Contributor Author

Might be handled soon in 0.5 emissary-ingress/emissary#1358

@ukclivecox
Copy link
Contributor Author

Ambassador has this now emissary-ingress/emissary#1127

@ukclivecox ukclivecox self-assigned this May 29, 2019
@ukclivecox
Copy link
Contributor Author

Config added by default now in master with latest seldon-operator

agrski pushed a commit that referenced this issue Dec 2, 2022
* move mlserver to grpc control plane

* move to mlserver 1.1

* tidy up unload all models logic

* add test

* only unload models that are READY or LOADING

* fix test

* k8s yaml files
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

2 participants