GnuTLS recv error (-9): A TLS packet with unexpected length was received

TKLBAM backup creation ends with the message: "Could not set backup end record GnuTLS recv error (-9): A TLS packet with unexpected length was received". This happens from both webmin and via SSH.

The backup is created but the TurnkeyLinux hub keeps showing "Backup in progress".
This happens when running a full backup.
The restore of this backup doesn't fully process.

So I am unable to make the full backup for now.

Anybody info on how to deal with this?

Replies (4)

That's a new one!

Reply 1 2014-06-08

I haven't encountered that issue before. A quick google suggests that its something to do with an untrusted SSL cert (although I can't be sure...). TBH I'm not sure quite what to do about it.

Could you retry now and let me know if it is still happening...

Proxy? NSA man-in-the-middle attack?

Reply 2 2014-06-10

First of all, having the full traceback would be more helpful than an excerpt. From the sound of it it seems like you might not have a direct connection to the AWS servers (or tot he Hub). Maybe something is intercepting the connection (e.g, a proxy on your network?)

More information required

Reply 3 2014-06-11

Great to have more information. The full error message (E.g., python traceback) would help pinpoint the issue. Also, if you add my SSH public key to the authorized keys on your server and send me the IP privately by email (e.g., liraz AT turnkeylinux.org) I could try logging in sometime run a few tests.

cat >> /root/.ssh/authorized_keys << EOF
ssh-rsa AAAAB3NzaC1yc2EAAAABIwAAAIEAvp+4FpjKlv1ngsdeiQtX8zMvQMkuJDwZSCHpFdm2IY20NmOhF0LY6dKRzQ+89pJ2MUYZYtotN1SmMk1ndUmHssQIRrmKKWdwnDzDUISTDB5iEQIg8JcPxwu6+uJnLrZvfNrx/fsMoRwRR3S9bHcKi9pxQT9T4Jbt+Gt6ewtuLAE= liraz
EOF

Only with full backup

Reply 4 2014-06-11

To get insight I cleared the backups from the hub and started with a new full backup. The message apears with this full backup. I tested restoring that full backup to my test VPS and that works. The subsequent incremental backup doesn't give this message! It only happens with the full backup. Both in SSH and Webmin. Since the backup works it is fine with me for now. I have 2 VPS linked to one hub account, one for live and one for test. On the test VPS I restore (full) backups from the live. The amount of data in a backup is 7 GB (15GB originally).