Re: [rdiff-backup-users] Client dying "randomly" on 1.1.16

[Armando M. Baratti]

Oliver Hookins escreveu:

We run rdiff-backup with standard priority, but I think for whatever reason

it is getting held up on large files while lots of I/O is taking place. I
grabbed a tcpdump of the traffic and it looks like there is normal SSH
traffic until it hits whatever snag, then there are no packets for two
hours.

Then the connecting machine sends a TCP keep-alive, to which the machine
being backed up sends a TCP reset. So I think maybe my keep-alive settings
need some tweaking.

What sort of settings do people use in their ssh configs for BatchMode,
ServerAliveCountMax, and TCPKeepAlive?

  

I've had the same behaviour as described above.
I've done a first rsync of my target (about 1.7 GB worth of data) on my backup machine. That takes about 20 min.
Then I've done a rdiff-backup with "--force" option just to build the rdiff-backup-data directory.
That takes  07:43h (no typo, more than 7 hours) to do that.
During most of this time the connection remained idle (I don't know why).

The only measure to maintain the connection alive was to set  "ServerAliveInterval 120" on ssh client side (/etc/ssh/ssh_config).
(I've also set "KeepAlive" to "no"  on the server side, so the connection would no be aborted if the client was not responding, but this alone could not avoid the connection break, and I've not returned the option when testing "ServerAliveInterval 120" on the client side).

I'm using version 1.0.5 of rdiff-backup (on both sides).
The CPU and memory consumption are very low, and the same with respect with disc activity.

I was blaming this to the difference on the versions of OpenSSH between the machines (4.3p2 on the backup, 3.5p1 on the target machine), but I'm not convinced of this anymore.

--
Armando M. Baratti

LARC
  /Pesquisa de Marketing Ltda
Alameda dos Maracatins, 161 - Indianópolis
CEP 04089-010 - São Paulo - SP
Tel.: (11) 5051-3133 - Fax: (11) 5052-5955
http://www.larc.com.br/