7.0.13's backup issue

Hello,
I’m trying to perform a straightforward database backup from the command line as a root user:
BACKUP DATABASE test TO ‘./’;
and I’m getting the following error message:

I used version 7.0.13 and it is on-premise. Not Helios.
Why can’t I get a local backup?

memsql> backup database monitoring to “/mnt/backup”;
ERROR 1706 (HY000): Feature ‘Writing a backup to the local file system when running MemSQL as a service’ is not supported by MemSQL.

Do I missing some things?

Please advise,
Thanks

Well, it looks like you found a bug in the error message. It should not be mentioning “running MemSQL as a service” when you’re running self-hosted. I’ll file a bug for that.

Was the directory writeable to the user running the memsqld process? Maybe there was an error, like that, but the error message was bad.

What does select @@local_file_system_access_restricted return?

The configuration of the cluster is as follows.
3hosts(ma+2leave)

Then, should all nodes have the same backup path?
What does that value mean?
If I did the right backup, should that be the number of nodes?

memsql> select @@local_file_system_access_restricted
    -> ;
+---------------------------------------+
| @@local_file_system_access_restricted |
+---------------------------------------+
|                                     0 |
+---------------------------------------+
1 row in set (0.01 sec)

thanks…

local_file_system_access_restricted of 0 is expected. That output is coming from the same cluster your getting the “Writing a backup to the local file system” error?

Can you send a cluster report to bug-report@memsql.com ?

Resolved.
The license that was applied at that time was a free license.
I’m not sure if the license does not support backup.

Changing the license to an enterprise resolves the problem.

thanks.
Ah, Do you need a report?

I wonder if all nodes should have the same backup path set to nfs?

BACKUP is supported in free edition. Send over the report if you get a chance.

The best way to do a backup is to NFS or a blob store (S3). Otherwise, the backup will write itself out to the local disk on each leaf node and you’ll have to manually move the files off the cluster yourself.