Backup issues: large TAR halts backup

I’ve been having this problem for a few weeks and I’m struggling to figure out what to do.

I have a scheduled full backup over SSH. When the VPS has been freshly restarted it usually works without issue, but after its been running for a few days backups will not complete. When this happens no log is saved in Backup Logs.

If I run the backup manually I can see that the first few virtual servers backup, TAR, and transfer successful. But when it gets to a larger virtual server it seems to fail during TAR. At this point I can see the whole backup and TAR process has been exited.

I suspect it could be a RAM issue, but I haven’t been able to locate any OOMs for TAR or in logs. The VPS has 1GB of RAM and 500MB of VSwap (its OpenVZ).

What should I do next?


As you suspected, it could very well be a resource issue; whether that be CPU, RAM or both. I’d suggest monitoring the task via “top” to see how much RAM and/or CPU is used while it’s running, leading up to the failure.

Assuming it’s a resource issue, either upgrade the VPS to a more suitable configuration, or perhaps schedule services like Apache, Postfix and other potentially resource intensive services to be temporarily shutdown during the backup window. This would free up resources and may aid in successfully completing the backups.

*** Keep in mind, this will bring down the availability of the services shutdown. ***

If you require any further assistance, feel free to post to this thread or drop me an email.

Best Regards, Peter Knowles TPN Solutions