6386 shutdown: I CAN BELIEVE AT&T was really this SMART!

Dennis S. Breckenridge root at nebulus.UUCP
Fri Jun 23 14:34:36 AEST 1989


In article <2294 at drilex.UUCP>, dricejb at drilex.UUCP (Craig Jackson drilex1) writes:
> In article <12044 at bloom-beacon.MIT.EDU> jfc at athena.mit.edu (John F Carr) writes:
> >In article <14401 at bfmny0.UUCP> tneff at bfmny0.UUCP (Tom Neff) writes:
> Personally, I've always thought that update(8) was a hack, ever since I
> first saw it in Version 7.  To think that someone would write an operating
> system that couldn't even keep its on-disk data structures consistent...
> 
Going back to V7 days when a file system corrupted itself you had to 
spend many hours on the console. This is reduced with the development
of several tools to allow you to repair a broken file system. With the 
release of file system hardening this was reduced even more. What is the 
problem with sync today and sync of V7 days. V7 sync forced a flush of the
system pool to disk, SYSV sync tags dirty buffers for a next scheduled 
update. 
 Consider for a moment several users updating the super-block and buffer
pools asyncronously. The system would spend a great deal of time on disk
and the user runtime would starve. By tagging buffers for update to happen
at the next AUTOUP time relieves the kernel from actually forcing a disk
write to complete before the call completes. This allows several buffers
to get written to before the actual memory to disk write happens. 
 My question really is how else would you do it effectively. I read many
complaints about how Eh-TNT did this but what I do not read is a 
proactive solution to the problem. Buy the source and fix it, send the 
diffs to Eh-TNT and see if you can add your name the list of thousands
that worked on it. 
 I believe that a company the size of AT&T that releases its product 
in source format for a small sub-license fee is doing the right thing.
Try that with the other vendors out there. Unix is portable, name 
some others with it's functionality and general availability. 

There is GNU...




-- 
==============================================================================
"A mind is a terrible thing to       MAIL:   Dennis S. Breckenridge
waste!"                                      206 Poyntz Ave
					     North York, Ontario M2N1J6
					     (416) 733-1696
UUCP: uunet!attcan!nebulus!dennis    ICBM:   79 28 05 W / 43 45 01 N
					     50 megatons should do!
==============================================================================



More information about the Comp.unix.wizards mailing list