Our current environment has a LAN infrastructure that is quite overloaded, and at peak times can bottleneck. In a lot of cases this can completely cut off connectivity for short periods of time for some hosts. Now our back-up PC runs during the early hours of the morning to prevent disruption. However, all of our other back-ups are spaced throughout the night too.

As back-ups sometime fails due to the above, we require something along the lines of a tick-box; resulting in automatic retries when back-ups fail. I'm not sure if such a feature exists though, as I've looked around multiple forms and done keyword searches to no prevail.

Appreciate any help offered. This is only a temporary measure we hope, as we'll be upgrading the infrastructure at some point.



2 Answers 2


Note at first: If there are Network problems they should be solved. usually one Problem will cause the next... (e.g. in your case: backup times, reliability of backups done)

However you may tune the network settings (which is more linux tcp topic, rathen than BackupPC)

Suggestion: You should be able to tune your system TCP network settings. from your description, the TCP connection brakes.

Background: - TCP connections will fail if the Connection is recognized to be dead. - The (Linux) TCP Stack has parameters to set Number of retries and Time between retries, Number of keep alive messages and time between them.

A solution for may be using wireshark to detect the kind of failure an to maniulate the TCP Settings on Server side and client side.

e.g. if missing keep alive packages causing the breach of the TCP Connection:

  • Try to set the keep alive time lower (tcp_keepalive_time)
  • set the number of keep alive packages before breaking the connections higher (tcp_keepalive_probes)
  • increase the timeout (tcp_keepalive_intvl)

Source: http://www.tldp.org/HOWTO/TCP-Keepalive-HOWTO/usingkeepalive.html


For full backups, if you adjust BackupPC settings, you can specify a Max age (e.g., 2 hours), and then a minimum number of backups to have. You can also schedule when backup attempts are created. If the software already has enough backups, then it'll skip the backups, because it already has the minimum and existing backups are within the desired age. But if a prior backup failed, it might not have enough backups to have the expected minimum, so then it will try again to create a backup. Then you can run a custom rotate script that moves/hides some of the backup data, to control how many backups are visible.

For incrementals, you could just make another attempt. If both attempts are successful, then the later incremental is likely to use few resources (disk space, network bandwidth).

  • Caution: I've just been learning this program in the last few days. You should consider my answer to be possible, but please verify before you count on its accuracy.
    – TOOGAM
    Nov 8, 2015 at 11:54

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .