WIP [vtctld] backup status

[ajm188]

Description

Demo

Note that for this, I made a quick change to builtinbackupengine's GetBackupStatus method to return a random status, via:

diff --git a/go/vt/mysqlctl/builtinbackupengine.go b/go/vt/mysqlctl/builtinbackupengine.go
index 3f88b5eecd..4888eb6165 100644
--- a/go/vt/mysqlctl/builtinbackupengine.go
+++ b/go/vt/mysqlctl/builtinbackupengine.go
@@ -23,6 +23,7 @@
 	"flag"
 	"fmt"
 	"io"
+	"math/rand"
 	"os"
 	"path"
 	"sync"
@@ -463,7 +464,16 @@ func (be *BuiltinBackupEngine) backupFile(ctx context.Context, params BackupPara
 // This is currently not implemented for builtinbackupengine, so we always
 // return UNKNOWN.
 func (be *BuiltinBackupEngine) GetBackupStatus(ctx context.Context, bh backupstorage.BackupHandle, mfestBytes []byte) (mysqlctlpb.BackupInfo_Status, error) {
-	return mysqlctlpb.BackupInfo_UNKNOWN, nil
+	// pick a random value from the map, only works because there are no gaps in the values
+	i := rand.Intn(len(mysqlctlpb.BackupInfo_Status_name))
+	status := int32(i)
+	if _, ok := mysqlctlpb.BackupInfo_Status_name[status]; !ok {
+		status = int32(mysqlctlpb.BackupInfo_UNKNOWN)
+	}
+
+	return mysqlctlpb.BackupInfo_Status(status), nil
+
+	// return mysqlctlpb.BackupInfo_UNKNOWN, nil
 }
 
 // ExecuteRestore restores from a backup. If the restore is successful

Anyway, the results, showing limiting and details (more to show that the backend impl is correct than anything about the CLI usability -- i am going to change the way we marshal json to convert enums from ints to their actual names separately):

❯ vtctlclient -server ":15999" Backup zone1-101
❯ vtctlclient -server ":15999" Backup zone1-102
❯ vtctldclient --server ":15999" GetBackups commerce/0
2021-06-15.003301.zone1-0000000101
2021-06-15.003310.zone1-0000000102
❯ vtctldclient --server ":15999" GetBackups commerce/0 -l1
2021-06-15.003310.zone1-0000000102
❯ vtctldclient --server ":15999" GetBackups commerce/0 --detailed
{
  "backups": [
    {
      "name": "2021-06-15.003301.zone1-0000000101",
      "directory": "commerce/0",
      "keyspace": "commerce",
      "shard": "0",
      "tablet_alias": {
        "cell": "zone1",
        "uid": 101
      },
      "time": {
        "seconds": "1623717181",
        "nanoseconds": 0
      },
      "engine": "builtin",
      "status": 2
    },
    {
      "name": "2021-06-15.003310.zone1-0000000102",
      "directory": "commerce/0",
      "keyspace": "commerce",
      "shard": "0",
      "tablet_alias": {
        "cell": "zone1",
        "uid": 102
      },
      "time": {
        "seconds": "1623717190",
        "nanoseconds": 0
      },
      "engine": "builtin",
      "status": 1
    }
  ]
}

Related Issue(s)

Checklist

• Tests were added or are not required
• Documentation was added or is not required

Deployment Notes

[ajm188]

Note: I've only implemented this for filebackupstorage and s3backupstorage for now.

[ajm188]

very small method extraction to let me reuse the error codes, but specifying a particular engine name rather than "whatever value the flag pointer has"

[github-actions]

This PR is being marked as stale because it has been open for 30 days with no activity. To rectify, you may do any of the following:

• Push additional commits to the associated branch.
• Remove the stale label.
• Add a comment indicating why it is not stale.

If no action is taken within 7 days, this PR will be closed.

[github-actions]

This PR was closed because it has been stale for 7 days with no activity.