On Wed, 31 Oct 2007, Ralf Gross wrote: >> I have pretty big full backup (3,2TB) for which I need around 12 LTO-2 tapes >> (with hw compression on). I use spooling with max size of 215GB, and >> typically spooling (of max size) lasts 2h50min, while despooling lasts >> 2h40min. Is this normal that spooling lasts longer than actual writing to >> the tapes? >> >> I would drop the spooling part, but I'm worried about tape shoe-shining. >> (If only bacula can do despooling and start spooling the next chunk of the >> same job at the same time.) > > Yes, even a simple FIFO with couple of GB or less would be > sufficient to avoid shoe-shining. Bacula's spooling has other > advantages than only avoid shoe-shining, but in our case it's not the > perfect solution.
have you benchmarked your spool partition with 1 (and more) simultaneous operations going on? I find that a 4-disk Raid0 stripeset has blistering performance for spooling/despooling, but performance falls off rapidly as you add more spooling/despooling operations - mainly due to head seeking. The answers appear to lay along the lines of: 1: More spindles - more R/W heads helps spread the load The problem is that suitable controllers can be pricey. or 2: Solid state disks - low-to-no seektimes. These are expensive (http://news.zdnet.co.uk/hardware/0,1000000091,39287193,00.htm), but prices are falling fast. I don't know how long they'd last in this kind of application. or 3: Both the above... ------------------------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Still grepping through log files to find problems? Stop. Now Search log events and configuration files using AJAX and a browser. Download your FREE copy of Splunk now >> http://get.splunk.com/ _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
