-
Notifications
You must be signed in to change notification settings - Fork 36
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
add exception to egress policy for port 8080 #924
Conversation
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.
This network policy is too permissive, as it grants wide IP range access to all workloads in a customer namespace. We need to carefully sandbox the network for customer-controlled workloads. I think it would be OK to apply a policy like this on the other network policy that applies only to the ai-proxy
9407c16
to
9ec818b
Compare
Yes thanks. I missed that specific policy. I just updated it and it should be good now, hopefully. |
"namespaceSelector": { | ||
"matchLabels": { | ||
"kubernetes.io/metadata.name": "tembo-ai" | ||
} | ||
}, | ||
"podSelector": { | ||
"matchLabels": { | ||
"app.kubernetes.io/name": "tembo-ai-gateway" | ||
} |
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.
Does removing this block break the path to connect via cluster IP within the same cluster to the tembo-ai services?
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.
It does. I can add it back if needed to allow for both
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.
I am not sure that all these statements work together in a single network policy, it might have to be separate network policies. I would try applying kind of configuration this manually in dev by disabling the reconcile on an instance, then apply this configuration, then see if it works. I think there is a pretty good chance this does not work because I recall there might be issues with specifying both ports and / or IP ranges in the same policy as namespace-selector based policies.
c6beebe
to
6fcf307
Compare
Ah yeah that is correct. I have created a new policy with what worked when testing. |
There needs to be an exception to the
allow-system-egress
policy to allow for access to new Internal load balancer for the tembo-ai gateway.We should allow access to
10.0.0.0/8
on port8080
. This will allow theai-proxy
pods access to the load balancer for the inference-gateway on all clusters.