On Tuesday 05 December 2006 13:50, Alan Brown wrote: > > Kern, > > Is there enough flexibility on the code to assign greater or lesser > priority to spooling/despooling processes?
Currently Bacula never fiddles with priorities, if for no other reason that it is very system dependent, and I try as best possible to avoid system dependent code. I haven't researched the problem, but it seems to me the first thing to do would be to see if it is possible to give different threads different priorities. If and only if it is possible, then a Feature Request would be needed. > > With current readily available tape technology (LTO2) running at speads > equal to most disk drives and newer technology (LTO3) running > significantly faster than this, there is probably some advantage in > allowing despooling processes to take higher priority over spooling ones > > (especially given there may be 5-10 processes spooling to disk > simultaneously at times) > > The only way I'm able to keep up with the tape drive and spool to disk > simultaneously at the moment is to run a 4-disk stripe set in order to > increase the spool area throughput speed (next step would be a dedicated > external SATA array) > > Even then, sometimes the tape starts shoeshining.... This problem is probably better solved by implementing a FIFO multiple buffering write scheme in the Storage daemon. I have always planned to do so, but it has never gotten even close to the top of my priorities. ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys - and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
