On Fri, Oct 01, 2010 at 08:07:27PM +0200, Sylvain MAURIN wrote: > Why didn't you check the amount of the IO ? FS bottlenecks, > particulary BackupPC heavy inode seeking, are *never* a result > of disk bandwith. > > Your choice of RAID 5, is really a pain in IO domain > if u don't take a lot of time in striping optimization. I understood that. This is only to backup 4 desktops and 3 remote servers.
> My advice : do an iotop/iostat -mtx 3 and find out > where is your problem. I am ready to bet that you > rarely have large sequential access if you backup many > desktop systems and around 2Mb you are closing 500 IO/s > that most RAID 5 Subsytems can provide. I'm getting about 1500 IO/s throughput at the moment. Sometimes higher Device: rrqm/s wrqm/s r/s w/s rMB/s wMB/s avgrq-sz avgqu-sz await svctm %util md3 0.00 0.00 472.00 1954.67 3.67 7.64 9.54 0.00 0.00 0.00 0.00 The biggest change is that the second full backup performed after turning off compression is 5 times faster (118 m) than the previous 5 backups (4 w/compression, 1 w/o compression) (450 - 700 m each). I'm hoping that that means rsync is doing it's job and unchanged files are not being transferred now. -- Wayne Walker [email protected] (512) 633-8076 Senior Consultant Solid Constructs, LLC ------------------------------------------------------------------------------ Start uncovering the many advantages of virtual appliances and start using them to simplify application deployment and accelerate your shift to cloud computing. http://p.sf.net/sfu/novell-sfdev2dev _______________________________________________ BackupPC-users mailing list [email protected] List: https://lists.sourceforge.net/lists/listinfo/backuppc-users Wiki: http://backuppc.wiki.sourceforge.net Project: http://backuppc.sourceforge.net/
