Handle more than about 64 TLS Knative Routes

[evankanderson]

(update from @dprotaso) See this comment for the current limit: #317 (comment)

We add a Listener (and a certificate reference) per Knative Route (KIngress). Each one is around 300 bytes, which means that we can probably fit around 3000 Routes into one Gateway before we exceed etcd storage limits.

According to some of the gateway maintainers, we should be able to split these across multilpe Gateways that all share the same IP address (so that users can still get that nice wildcard DNS mapping).

This issue tracks packing Listeners into multiple Gateways, it's extracted from[ this PR comment

[youngnick]

Wearing my Gateway API maintainer hat, it's absolutely intended that implementations may merge Gateways together into a single data plane config, but we don't mandate it. So the sort of sharding you're describing is definitely possible.

The exact details of how it will work may vary with the underlying data path implementation though.

[dprotaso]

Interestingly enough we've seen Routes take longer to become 'Ready=True` as the number of increases. The root cause seems to be the underlying ingress implementation

ie. see contour HTTProxy's timings: projectcontour/contour#4058

I don't think we've done similar tests with TLS routes where gateways & listeners increase

[github-actions]

This issue is stale because it has been open for 90 days with no
activity. It will automatically close after 30 more days of
inactivity. Reopen the issue with /reopen. Mark the issue as
fresh by adding the comment /remove-lifecycle stale.

[knative-prow-robot]

This issue or pull request is stale because it has been open for 90 days with no activity.

This bot triages issues and PRs according to the following rules:

• After 90d of inactivity, lifecycle/stale is applied
• After 30d of inactivity since lifecycle/stale was applied, the issue is closed

You can:

• Mark this issue or PR as fresh with /remove-lifecycle rotten
• Close this issue or PR with /close

/lifecycle stale

[github-actions]

This issue is stale because it has been open for 90 days with no
activity. It will automatically close after 30 more days of
inactivity. Reopen the issue with /reopen. Mark the issue as
fresh by adding the comment /remove-lifecycle stale.

[ReToCode]

/remove-lifecycle stale

[github-actions]

This issue is stale because it has been open for 90 days with no
activity. It will automatically close after 30 more days of
inactivity. Reopen the issue with /reopen. Mark the issue as
fresh by adding the comment /remove-lifecycle stale.

[ReToCode]

/remove-lifecycle stale

[dprotaso]

I've created a GEP issue to help address this - kubernetes-sigs/gateway-api#1713

Note the Gateway API folks have a limit on the # of listeners per Gateway and it's 64