From: Kern S. <ke...@si...> - 2002-11-29 11:41:00
|
Hello, I haven't seen any responses to this, so I'll throw in my 2 cents hoping that I don't discourage anyone else from responding too. In general Unix is a really nice operating system with well documented simple system calls. Over the years, it has slowly evolved (threads, ...). However, there are two areas in which I work where Unix is really a MESS. They are USB programming and tape handling. In the case of USB, the API is quite extensive but unfortunately not standard from one Unix to another. In addition, the USB specification is really complicated (no fault of Unix). The other area that interests us here is tape handling. Here the API is almost non-existent, and the few ioctl() calls that do exist don't give the programmer the information he needs (such as min/max block sizes, whether an EOT was reached or a real error occurred, ...). In addition, the stupid ioctl()s that do exist are for the most part not system independent and are very poorly documented to the extent that the Linux implementors didn't know what some ioctl() fields are really supposed to do! In other words, it is a mess. As you can see, this really frustrates me. Just the same, if anyone has a tape drive that doesn't work with Bacula, I am very willing to spend a fair amount of effort to try to provide support for it. In fact, the current Bacula tape driver is quite adaptable. If you understand the behavior of your drive, it is very probable that by using the correct set of Bacula records in the Device resource, it will work. Just the same, there is certainly more I can do to support these devices given access to such a device or proper feedback. Best regards, Kern On Mon, 2002-11-25 at 03:51, Adrian Close wrote: > Hi all, > > Sorry if this is a little off-topic, but I figure the readership of this > list is more likely than most to have a clue about my question. Which is > along these lines: > > I have a Seagate ATAPI tape drive (labelled as a STT320000A) that I'd like > to use with Bacula (on RH8.0). Unfortunately, it fails the btape tests, > with respect to tape positioning. I'm guessing this is because of driver > issues. > > So, just as a first pass, does this sound familiar to anyone? Pointers to > where I should be looking? I'm not afraid of rebuilding kernels or even > hacking source if that should be required - just looking for expressions > of solidarity (or alternatively someone to stop me wasting my time!). > > Incidentally, I have a SCSI version of this same drive which works just > fine... > > Thanks, > > Adrian Close email: ac...@sa... > NOC Analyst mobile: +61 411 828 823 > Saise phone: +61 3 8635 1738 > 10/50 Queen St, Melbourne, VIC, 3000, AU fax: +61 3 9620 4887 > > > > ------------------------------------------------------- > This sf.net email is sponsored by:ThinkGeek > Welcome to geek heaven. > http://thinkgeek.com/sf > _______________________________________________ > Bacula-users mailing list > Bac...@li... > https://lists.sourceforge.net/lists/listinfo/bacula-users |