Add vtctldclient missing cmds

[mattlord]

Description

This implements client commands in vtctldclient that were never ported during the client migration. These are things that were identified while moving all endtoend tests to use vtctldclient exclusively in #17441.

We will (after merging in main once #17441 is merged) also then replace all remaining vtctlclient usage in the e2e tests with usage of these new vtctldclient commands.

Related Issue(s)

• Follow-up to: Move all e2e tests to vtctldclient #17441

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

Attention: Patch coverage is 11.18644% with 262 lines in your changes missing coverage. Please review.

Project coverage is 67.61%. Comparing base (059d01a) to head (9e87209).
Report is 10 commits behind head on main.

Files with missing lines	Patch %	Lines
go/vt/vtctl/grpcvtctldserver/server.go	7.00%	93 Missing ⚠️
go/cmd/vtctldclient/command/schema.go	18.44%	84 Missing ⚠️
go/cmd/vtctldclient/command/permissions.go	8.51%	43 Missing ⚠️
go/cmd/vtctldclient/command/topology.go	9.67%	28 Missing ⚠️
go/vt/vtctl/grpcvtctldclient/client_gen.go	0.00%	10 Missing ⚠️
go/vt/vtctl/localvtctldclient/client_gen.go	0.00%	4 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17442      +/-   ##
==========================================
- Coverage   67.68%   67.61%   -0.08%     
==========================================
  Files        1583     1585       +2     
  Lines      254321   254596     +275     
==========================================
+ Hits       172131   172135       +4     
- Misses      82190    82461     +271     

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

[dbussink]

Does this not contain any results?

[mattlord]

Yeah, it returns an error if there's a validation failure:

vitess/go/vt/vtctl/vtctl.go

Lines 3259 to 3269 in 76abae4

	func commandValidatePermissionsKeyspace(ctx context.Context, wr *wrangler.Wrangler, subFlags *pflag.FlagSet, args []string) error {
	if err := subFlags.Parse(args); err != nil {
	return err
	}
	if subFlags.NArg() != 1 {
	return fmt.Errorf("the <keyspace name> argument is required for the ValidatePermissionsKeyspace command")
	}
	keyspace := subFlags.Arg(0)
	return wr.ValidatePermissionsKeyspace(ctx, keyspace)
	}

We could change that, of course, but here I'm primarily just porting these missing commands over.

[shlomi-noach]

We have this as common practice to sometimes have empty responses, where it's "error or nothing" and with the potential of adding proto fields in the future without breaking gRPC signature.

[dbussink]

Guess this is more a question for a separate change perhaps. But given we have various inputs with different shards, wouldn't we want to reflect that in the output similar to the next command listed here?

[shlomi-noach]

Valid point, worth extended discussion.

[mattlord]

"We" here would be whomever uses and consumes this command (certainly not me to date). This PR is about migrating the missing but still used/needed commands from the old client to the new, where maintaining "drop in compatibility" is the only concern. Why are we discussing this here? Is there some feature gap or request? If so, we should open a feature request that lays it out and someone can pick that up at a later date.

I'm actually leaning towards removing the endtoend test code for it (ValidatePermissions*, as I did with CopySchemaShard) since I cannot find anything that actually uses it today (in vitess, vitess-operator, etc):

go/test/endtoend/sharded/sharded_keyspace_test.go:      err = clusterInstance.VtctlclientProcess.ExecuteCommand("ValidatePermissionsShard", fmt.Sprintf("%s/%s", keyspaceName, shard1.Name))
go/test/endtoend/sharded/sharded_keyspace_test.go:      err = clusterInstance.VtctlclientProcess.ExecuteCommand("ValidatePermissionsKeyspace", keyspaceName)
go/vt/vtctl/vtctl.go:                           name:   "ValidatePermissionsShard",
go/vt/vtctl/vtctl.go:                           method: commandValidatePermissionsShard,
go/vt/vtctl/vtctl.go:                           name:   "ValidatePermissionsKeyspace",
go/vt/vtctl/vtctl.go:                           method: commandValidatePermissionsKeyspace,
go/vt/vtctl/vtctl.go:func commandValidatePermissionsShard(ctx context.Context, wr *wrangler.Wrangler, subFlags *pflag.FlagSet, args []string) error {
go/vt/vtctl/vtctl.go:           return fmt.Errorf("the <keyspace/shard> argument is required for the ValidatePermissionsShard command")
go/vt/vtctl/vtctl.go:   return wr.ValidatePermissionsShard(ctx, keyspace, shard)
go/vt/vtctl/vtctl.go:func commandValidatePermissionsKeyspace(ctx context.Context, wr *wrangler.Wrangler, subFlags *pflag.FlagSet, args []string) error {
go/vt/vtctl/vtctl.go:           return fmt.Errorf("the <keyspace name> argument is required for the ValidatePermissionsKeyspace command")
go/vt/vtctl/vtctl.go:   return wr.ValidatePermissionsKeyspace(ctx, keyspace)
go/vt/vtctld/vtctld.go: actionRepo.RegisterKeyspaceAction("ValidatePermissionsKeyspace",
go/vt/vtctld/vtctld.go:                 return "", wr.ValidatePermissionsKeyspace(ctx, keyspace)
go/vt/vtctld/vtctld.go: actionRepo.RegisterShardAction("ValidatePermissionsShard",
go/vt/vtctld/vtctld.go:                 return "", wr.ValidatePermissionsShard(ctx, keyspace, shard)
go/vt/wrangler/permissions.go:// ValidatePermissionsShard validates all the permissions are the same
go/vt/wrangler/permissions.go:func (wr *Wrangler) ValidatePermissionsShard(ctx context.Context, keyspace, shard string) error {
go/vt/wrangler/permissions.go:// ValidatePermissionsKeyspace validates all the permissions are the same
go/vt/wrangler/permissions.go:func (wr *Wrangler) ValidatePermissionsKeyspace(ctx context.Context, keyspace string) error {
go/vt/wrangler/permissions.go:          return wr.ValidatePermissionsShard(ctx, keyspace, shards[0])
go/vt/wrangler/testlib/permissions_test.go:     // run ValidatePermissionsKeyspace, this should work
go/vt/wrangler/testlib/permissions_test.go:     if err := vp.Run([]string{"ValidatePermissionsKeyspace", primary.Tablet.Keyspace}); err != nil {
go/vt/wrangler/testlib/permissions_test.go:             t.Fatalf("ValidatePermissionsKeyspace failed: %v", err)
go/vt/wrangler/testlib/permissions_test.go:     // run ValidatePermissionsKeyspace again, this should now fail
go/vt/wrangler/testlib/permissions_test.go:     if err := vp.Run([]string{"ValidatePermissionsKeyspace", primary.Tablet.Keyspace}); err == nil || !strings.Contains(err.Error(), "has an extra user") {
go/vt/wrangler/testlib/permissions_test.go:             t.Fatalf("ValidatePermissionsKeyspace has unexpected err: %v", err)

So rather than spending time discussing how we could improve it... I think we should kill it off during the client transition as it no longer seems to serve a valid/used purpose as a client command. Unless you for some reason feel this is needed (and should be improved at a later time)?

[mattlord]

To provide some context, the GetPermissions client command simply returns the contents of the mysql.user and mysql.db tables as a JSON doc (tmc -> mysqlctl):

❯ vtctldclient GetPermissions zone1-100
{
  "user_permissions": [
    {
      "host": "%",
      "user": "vt_repl",
      "password_checksum": "0",
      "privileges": {
        "Alter_priv": "N",
        "Alter_routine_priv": "N",
        "Create_priv": "N",
        "Create_role_priv": "N",
        "Create_routine_priv": "N",
        "Create_tablespace_priv": "N",
        "Create_tmp_table_priv": "N",
        "Create_user_priv": "N",
        "Create_view_priv": "N",
        "Delete_priv": "N",
        "Drop_priv": "N",
        "Drop_role_priv": "N",
        "Event_priv": "N",
        "Execute_priv": "N",
        "File_priv": "N",
        "Grant_priv": "N",
        "Index_priv": "N",
        "Insert_priv": "N",
        "Lock_tables_priv": "N",
        "Password_require_current": "",
        "Password_reuse_history": "",
        "Password_reuse_time": "",
        "Process_priv": "N",
        "References_priv": "N",
        "Reload_priv": "N",
        "Repl_client_priv": "N",
        "Repl_slave_priv": "Y",
        "Select_priv": "N",
        "Show_db_priv": "N",
        "Show_view_priv": "N",
        "Shutdown_priv": "N",
        "Super_priv": "N",
        "Trigger_priv": "N",
        "Update_priv": "N",
        "User_attributes": "",
        "account_locked": "N",
        "authentication_string": "",
        "max_connections": "0",
        "max_questions": "0",
        "max_updates": "0",
        "max_user_connections": "0",
        "password_expired": "N",
        "password_lifetime": "",
        "plugin": "mysql_native_password",
        "ssl_cipher": "",
        "ssl_type": "",
        "x509_issuer": "",
        "x509_subject": ""
      }
    },
...

The ValidatePermissions* commands then tells you if the output differs within a shard or a keyspace. I don't disagree that telling you what nodes differ and how would be useful (not in this PR, but later enhancements)... IF this is actually still useful in the broader sense today. I haven't yet seen anything to indicate that it's still used or still useful.

[mattlord]

In looking for uses of the ValidatePermissons commands (I found none) I did find uses of CopySchemaShard. So I think I will have to port that over.

[mattlord]

One last note... FWIW the functions used for the ValidatePermissions* commands do include the details in the error message:

vitess/go/vt/mysqlctl/tmutils/permissions.go

Line 164 in a57ae93

func diffPermissions(name, leftName string, left permissionList, rightName string, right permissionList, er concurrency.ErrorRecorder) {

So I'm going to put this comment thread to bed for now. 🙂