[comp.sys.nsc.32k] df reports "not a filesystem"

emc@unicus.UUCP (Eric M. Carroll) (05/24/87)

A strange thing began to happen after we installed our new Maxtor 3380
drives this weekend. We used 512B blocks and partitioned the disk
as follows:
	0s0:	16384 blocks	(/)	0s4:	64 blocks (boot)
	0s1:	526604 blocks	(/usr)	0s5:	16384	  (swap)
	0s2:	61440 blocks	(/u)	0s6:	2	  (partition table)
	0s3:	overlaps 0s1 and 0s2	0s7:	620876	  (disk)
The layout was / swap /usr /u, which is rather non-standard. The problem
is that df reports 
	(/dev/dsk/0): is not a file system
once /dev/dsk/0s1 is mounted on /usr. I looked at the strings in df, and
saw that it was using a ten wide field for the device name - thus the 
truncation. But why is it not a filesystem? mkfs worked ok, and
the Great Disk God fsck says nary a peep. In fact, we have been running
it as a filesystem for a day now, no trouble. We have no source, so I 
cannot divine what df considers to be an illegal file system.
The strings output shows that df is using a ld format in printing out the
number of blocks, killing our pet theory of shorts in use when longs
were needed. This is messing up my accounting, so if all else fails we
will have to rewrite df from scratch.

This is all on sysVr2v2 on an ICM3216 from NSC.
	
	Any help would be appreciated. Thanks.
-- 
	Eric Carroll 			Unicus Corporation, Toronto Ont.
	Eric.M.Carroll@Unicus.COM				  (Internet)
	{seismo!mnetor, utzoo!utgpu!utcsri}!unicus!Eric.M.Carroll (dumb UUCP)
	mnetor!unicus!Eric.M.Carroll@seismo.css.GOV		  (dumb ARPA)

joer@nscpdc.NSC.COM (G. Wiz) (05/29/87)

In article <570@unicus.UUCP>, emc@unicus.UUCP (Eric M. Carroll) writes:
> A strange thing began to happen after we installed our new Maxtor 3380
> drives this weekend. We used 512B blocks and partitioned the disk
> as follows:
> 	0s0:	16384 blocks	(/)	0s4:	64 blocks (boot)
> 	0s1:	526604 blocks	(/usr)	0s5:	16384	  (swap)
> 	0s2:	61440 blocks	(/u)	0s6:	2	  (partition table)
> 	0s3:	overlaps 0s1 and 0s2	0s7:	620876	  (disk)
> The layout was / swap /usr /u, which is rather non-standard. The problem
> is that df reports 
> 	(/dev/dsk/0): is not a file system
> But why is it not a filesystem? mkfs worked ok, and fsck says nary a peep.
> 
> This is all on sysVr2v2 on an ICM3216 from NSC.
> 	
> 	Any help would be appreciated. Thanks.
> -- 
> 	Eric Carroll 			Unicus Corporation, Toronto Ont.

Eric,

In order to help you, I need more information.  I have tried "df" on
many drives and the only time I see that message is when I specify a disk
partition that is not a file system.  (e.g. "df /dev/dsk/0s5").

What release of the software do you have?  
What is mounted at the time you issue the command?  
What parameters were passed to "df" when you got the above message?
What do you mean by "The layout was / swap /usr /u." ?

Note: the swap partition CANNOT be and IS NOT a file system.  If 
you have mounted /dev/dsk/0s5 as anything, using "df" will cause the 
above error message to be generated.  

A "mkfs" on /dev/dsk/0s5 will work but is a complete waste of time 
because the first time that Unix swaps anything out, the superblock 
on /dev/dsk/0s5 is no more!  The UNIX internal superblock is OK and "df" 
will not report an error until after /dev/dsk/0s5 is umounted, then 
again mounted.  At that time, "df" will generate the above error message.

If you have a new bug we will be glad to check it out.  I would also like
to recommend that you call ICM Product Support directly with your problems
as they may not be.  From Canada, use the number (503) 629-4414. 

An aside: With all that disk space, why not increase the swap partition?


-- 
/****************************************************************************** *                                                                             *
*  Joe Rawlings                         nsc!nscpdc!joer         or            * *  ICM Product Support                  tektronix!reed!nscpdc!joer            *
*  National Semiconductor Corp.         1-800-222-2433 (outside Oregon)       * *  Portland Development Center          or    629-4414 (inside Oregon)        *
*                                                                             * ******************************************************************************/