Some additional data points that might be of interest:
I performed a set of simple experiments using the following methods:
rdiff-backup
rsnapshot-ssh
rsnapshot-rsycn-no-compression-whole-file
rsync-avz-ssh
rsync-server-no-compression
rsync-server-no-compression-whole-file
I'll spare you the details, but these two performed nearly identically:
rsnapshot-rsync-no-compression-whole-file
rsync-server-no-compression-whole-file
And during backup execution, they both had minimal impact on the server
load. I have now switched over to using rsnapshot over an rsync daemon
for our high-volume production servers.
Cheers,
jason
Jason L. Buberel wrote:
And a good hunch at that. In fact, that was the first test I performed
- enable hot backup mode but do not invoke rdiff-backup. There is no
performance or load impact, indicating that the postgres hot-backup
feature is as well implemented as has been claimed :)
-jason
Austin Clements wrote:
Out of curiosity, what happens if you just put Postgres into hot
backup mode and don't run rdiff-backup? I've never personally
administered Postgres, but it seems that hot backup mode would cause
it to sync much more frequently, which could send your iowait time
through the roof.
Just a hunch.
_______________________________________________
rdiff-backup-users mailing list at [email protected]
http://lists.nongnu.org/mailman/listinfo/rdiff-backup-users
Wiki URL: http://rdiff-backup.solutionsfirst.com.au/index.php/RdiffBackupWiki
