On Wednesday 24 February 2010 17:03:58 Josh Fisher wrote: > On 2/24/2010 9:25 AM, Silver Salonen wrote: > > It's like assuming that the "ultimate" backup-devices are tapes. And as I > > don't think that way, it's so annoying these design decisions rely on > > somebody's (emotional/historical) opinion. > > > > The ultimate is a stream of bytes that makes up a Bacula volume. In this > way, a bacula volume is not media dependent, in that it doesn't matter > whether that volume is on tape, disk file, DVD, FIFO, etc. The design > decision separates backup function from I/O details.
The question was about whether and when to allow multiple volumes to be written on the same device concurrently. You were talking about what? > > What's the use of treating all the devices the same way anyway? Ease of > > programming? Even though it makes this part of the whole project so rigid? > > > > Bacula is following the Unix paradigm of "everything is a file", which > is to separate functionality from i/o details. Not everyone agrees, > perhaps, but it is a time tested method that has proven to work and is > trusted. Trust is of primary importance for any backup system. If volumes were files, there wouldn't be any need to limit them for devices which would be directories in that context. -- Silver ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ Bacula-users mailing list Bacula-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/bacula-users
