DEC R80 disk partitions (help wanted)

Scott Williamson scott at dice.UUCP
Thu Oct 25 22:24:19 AEST 1984


[..........]

We have a Vax 11/730 running 4.2bsd. Mass storage on this machine consists
of a Fujitsu Eagle and a DEC R80. My problem is trying to understand
how the disk partitions have been set up. (I took on the role of system
manager about 3 months after first encountering Unix.) Our root filesystem
is on the Eagle with swap space in the 'b' partitions of both disks. I
am having no problems with the Eagle but have encountered some difficulty
with the R80. The relevant entry in /etc/disktab is as follows:

	rb80|RB80|DEC R80 on 730 IDC:\
		:ty=winchester:ns#31:nt#14:nc#559:\
		:pa#15884:ba#8192:fa#1024:\
		:pb#33440:pc#242606:\
		:pg#82080:bg#4096:fg#512:\
		:ph#110143:bh#4096:fh#1024:

Looking at the source of the device driver (/sys/vaxuba/idc.c) there
seems to be agreement with the above.

	}, rb80_sizes[8] ={
		15884,	0,		/* A=cyl 0 thru 36 */
		33440,	37,		/* B=cyl 37 thru 114 */
		242606,	0,		/* C=cyl 0 thru 558 */
		0,	0,
		0,	0,
		0,	0,
		82080,	115,		/* G=cyl 115 thru 304 */
		110143,	305,		/* H=cyl 305 thru 558 */

However, using diskpart(8) to calculate default partition sizes for it
gives the following results:

	% diskpart -d rb80
	rb80:\
		:ty=winchester:ns#31:nt#14:nc#559:\
		:pa#15884:ba#8192:fa#1024:\
		:pb#33440:\
		:pc#242606:\
		:pd#15884:bd#8192:fd#1024:\
		:pe#55936:be#4096:fe#512:\
		:pf#120466:bf#4096:ff#1024:\
		:pg#192510:bg#4096:fg#1024:

It is possible that our /etc/disktab and device driver are the result
of a local hack since I do not have much idea of what my predecessor
did when he installed the file systems. At any rate, the partitions do
seem to be set up as in the first example ie. about 40Mb for partition g.
The problem is that we do not use partition h and I don't know why. As
far as I can gather, we have 55Mb of unused space on the disk. Is there a
problem with using the whole R80, or should I just go ahead and

	% newfs /dev/rb0h rb80

One thing I noticed when looking at /etc/disktab was that the entry for the
number of sectors in partition h (110143) is the only entry with an odd
number in it in the whole table ie. with block size 4Kb and fragment size
1Kb, it will not divide up evenly. Any help on this matter would be greatly
appreciated.

If anyone is still awake at this point, I have one more question to ask.
Where does the kernel learn what area of the disk to use as the secondary
swap space? Is it hard-wired into the system from the entry in the
configuration file ie. 

config		vmunix		root on XXXa swap on XXXb and XXXb

or is it read from the appropriate entry in /etc/fstab ie.

	/dev/XXXb::sw::

or both? What happens if the entries do not agree with each other?

Any replies should be sent by mail please.  Thanks in advance,

Scott Williamson.
{ ...!decvax!mcvax or ...!vax135!ukc!edcaad } !dice!scott



More information about the Comp.unix.wizards mailing list