Remove broken panic handler

[dbussink]

This handler would recover any arbitrary panic and make the test pass. This is a big anti-pattern since it means we can have many failing tests we don't know about and things are silently broken potentially.

Running against CI to test the fallout of this.

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

[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 67.46%. Comparing base (0fe256e) to head (3e175b7).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main   #17354   +/-   ##
=======================================
  Coverage   67.46%   67.46%           
=======================================
  Files        1581     1581           
  Lines      253934   253934           
=======================================
+ Hits       171308   171313    +5     
+ Misses      82626    82621    -5     

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

[dbussink]

@shlomi-noach @mattlord @rohit-nayak-ps This line was crashing this test, so we never actually ran it and it was always passing because of the now removed defer cluster.PanicHandler(nil) call. Looks like the test might now be failing, but it means it has likely been failing forever / for a really long time.

Any ideas on how to fix this?

[mattlord]

It's failing here:

func testTabletRecovery(t *testing.T, binlogServer *binLogServer, lookupTimeout, restoreKeyspaceName, shardName, expectedRows string) {
	recoveryTablet := clusterInstance.NewVttabletInstance("replica", 0, cell)
	launchRecoveryTablet(t, recoveryTablet, binlogServer, lookupTimeout, restoreKeyspaceName, shardName)

	sqlRes, err := recoveryTablet.VttabletProcess.QueryTablet(getCountID, keyspaceName, true)
	require.NoError(t, err)
	assert.Equal(t, expectedRows, sqlRes.Rows[0][0].String())

So my first guess would be that we need to wait for replication to catch up and for the expected result. Currently all it's doing is waiting for the tablet to become SERVING and then it queries it. Perhaps the tablet needs a second to replicate things. I would first just try adding a sleep of 30 seconds or something.

[shlomi-noach]

This tests the old-and-unsupported-and-actually-legacy Google Ripple binlog server based PITR.

• RFC: Backup/restore: remove legacy binlog-server based PITR code #16673
• https://vitess.io/docs/22.0/reference/features/recovery/#point-in-time-recovery-legacy-functionality-based-on-binlog-server

We should stop running this test altogether. Now is as good a time as ever.

[shlomi-noach]

@dbussink I've pushed a change to remove the test from running in shard 10. For now this should fix this PR. Later on (on a different PR) I will purge the entire codebase and tests.

[frouioui]

Aren't the tests supposed to fail anyway if there is a panic? Seems like cluster.PanicHandler should make the test fail if we recover a panic since the err wouldn't be nil:

vitess/go/test/endtoend/cluster/cluster_util.go

Lines 130 to 136 in 9b71606

	func PanicHandler(t testing.TB) {
	err := recover()
	if t == nil {
	return
	}
	require.Nilf(t, err, "panic occured in testcase %v", t.Name())
	}

[dbussink]

Aren't the tests supposed to fail anyway if there is a panic? Seems like cluster.PanicHandler should make the test fail if we recover a panic since the err wouldn't be nil:

It doesn't. In all the cases removed here, nil is passed in for t. So that means it does the recover() and then returns without failing the test.

So hence the test passes when this is used if the test panics.

[dbussink]

Backporting this too since it can be hiding legitimate test failures / problems and we don't want to have those crop up in release branches either then.