e2e: disable gRPC GOAWAY/too_many_pings keepalive errors in e2e CI

[timvaillancourt]

Description

This PR disables too-many-ping protection in gRPC when running e2e tests with test.go. This is achieved by setting GRPC_ARG_HTTP2_MAX_PINGS_WITHOUT_DATA to 0

Context from docs:

GRPC_ARG_HTTP2_MAX_PINGS_WITHOUT_DATA
This channel argument controls the maximum number of pings that can be sent when there is no data/header frame to be sent. gRPC Core will not continue sending pings if we run over the limit. Setting it to 0 allows sending pings without such a restriction. (Note that this is an unfortunate setting that does not agree with A8-client-side-keepalive.md. There should ideally be no such restriction on the keepalive ping and we plan to deprecate it in the future.)

and

Why am I receiving a GOAWAY with error code ENHANCE_YOUR_CALM?
A server sends a GOAWAY with ENHANCE_YOUR_CALM if the client sends too many misbehaving pings as described in A8-client-side-keepalive.md. Some scenarios where this can happen are -
if a server has GRPC_ARG_KEEPALIVE_PERMIT_WITHOUT_CALLS set to false while the client has set this to true resulting in keepalive pings being sent even when there is no call in flight.
if the client's GRPC_ARG_KEEPALIVE_TIME_MS setting is lower than the server's GRPC_ARG_HTTP2_MIN_RECV_PING_INTERVAL_WITHOUT_DATA_MS.

Today, many e2e tests connect/reconnect to the gRPC server too quickly and hit the GOAWAY/too_many_pings error. When this occurs, the victim test has to wait the GRPC_ARG_HTTP2_MIN_RECV_PING_INTERVAL_WITHOUT_DATA_MS default of 300000 milliseconds (5 minutes) doing nothing 😱. This protection makes sense in a normal world, but in an e2e test all running on localhost we don't really care

When tests hit this, we see this error and pause for 5min:

E1218 18:18:06.674435  32993 component.go:44] [transport] Client received GoAway with error code ENHANCE_YOUR_CALM and debug data equal to ASCII "too_many_pings".

Docs: https://github.com/grpc/grpc/blob/master/doc/keepalive.md

Related Issue(s)

Checklist

• "Backport to:" labels have been added if this change should be back-ported to release branches
• If this change is to be back-ported to previous releases, a justification is included in the PR description
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on CI?
• Documentation was added or is not required

Deployment Notes

AI Disclosure

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 69.90%. Comparing base (7a3acd5) to head (5a41620).
⚠️ Report is 6 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main   #19083   +/-   ##
=======================================
  Coverage   69.90%   69.90%           
=======================================
  Files        1612     1612           
  Lines      215817   215789   -28     
=======================================
- Hits       150865   150849   -16     
+ Misses      64952    64940   -12     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[mattlord]

@timvaillancourt IMO these logs are not helpful, at all, to the end user. Why don't we disable them entirely in Vitess rather than just in the tests?

[mattlord]

❤️

[timvaillancourt]

@timvaillancourt IMO these logs are not helpful, at all, to the end user. Why don't we disable them entirely in Vitess rather than just in the tests?

@mattlord unfortunately the gRPC library can't silence the logging unless you disable the protection entirely like we're doing here. I don't recommend we do that in a non-test environment

I think the built-in protection still makes sense in a non-e2e use case, because it's possible for the gRPC port to abused by attackers. Arguably this change shouldn't have been necessary if we used the gRPC client by-the-book, but it seems we often launch a single tmclient per-go-package in go/test/endtoend (via TestMain) that doesn't close it's connections properly - I believe this is the reason we really see this, and hypothetically if we were more careful we wouldn't see this

Anecdotally, the only time I've seen this error/wait as a Vitess user in a real deployment was when writing external automation that hits vtctld APIs, possibly without closing connections properly as well, or perhaps doing something too quickly (the details of the in-gRPC protection aren't too clear to me)