chris@umcp-cs.UUCP (Chris Torek) (11/10/86)
In article <131@houligan.UUCP> someone---we never received the original article, so I cannot say who---wrote: >a volume or set number, for implementing multivolume file systems (for >instance, Chris Torek's striping setup). Just to set the record straight, I have no striping setup. I mentioned one as a merit of a particular implementation (inode-based partitions) simply because it seemed to fit well with that implementation. (It seems to me that striping should be done either at the file system level or between file systems and device drivers; in either case, per-pack data seems not the proper place for striping information. [That it may *interact* with such information is another matter.] Putting striping data in the inode precludes striping the root file system, but given the importance of that file system, and the increased chance for chaos with multiplied mechanics, that seems not unreasonable.) -- In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690) UUCP: seismo!umcp-cs!chris CSNet: chris@umcp-cs ARPA: chris@mimsy.umd.edu