SUMMARY: inode blocks/cyl group (377) >= data blocks (224)

From: Adam Singer (
Date: Tue Jun 08 1999 - 23:02:11 CDT

Dear Sun Managers,

Thanks again for the speedy and very helpful responses. Thanks to:

Arthur Darren Dunham; Russ Poffenberger ; Ray Saddler ; Charlie
Mengler; Leif Ericksen; Brion Leary; Thomas Carter; David H. Brierley

Amongst the suggestions werre:

1. overlapping cyclinder (no)
2. greater than 4GB filesystem (never a problem before)
3. format.dat for the 9GB
4. But the correct answer is that I had it off cyclinder. I guess I
realized this but I *kept trying* various combinations. Still Russ
Poffenberger even excerpted the mkfs_ufs man page which I guess I
overlooked (sorry folks)!

Here are some excerpts that summarize the problem and solution better
than I could.

David Brierley: "This message usually means that the partition size is
just slightly larger than a multiple of 16. I notice that the output
of 'newfs' shows a size of 3136 cylinders. Does the output of the
format command agree with that? I suspect that the format command
will show that the partition size is 3137 cylinders and that is what
is causing your error."

Russ Poffenberger: "The mkfs_ufs man page explains the error..

     Warning: inode blocks/cyl group (grp) >= data blocks (num) in
          User request for inodes/byte (with the nbpi keyword)
          and the disk geometry results in a situation in which
          the last truncated cylinder group can not contain the
          correct number of data blocks; some disk space is

Arthur Darren Dunham: "To eliminate it, go back into 'format' and only
allocate partitions on whole cylinders."

thank all of you for your help

Adam Singer
> Hello,
> I am having some problems newfs'ing a 9GB disk on a Sun Ultra60
> running fully patched Solaris 2.6. I had had 9GB disks under
> DiskSuite with a couple of metadevices, but I had wanted to partition
> this disk just as a ufs filesystem. One of the slices was to be 5GB.
> Well everytime I ran newfs, no matter how I set the # of sectors from
> within format, I always got the following error:
> >Warning: inode blocks/cyl group (377) >= data blocks (224) in last
> > cylinder group. This implies 3590 sector(s) cannot be allocated.
> I have never seen this error before and had hoped someone could
> explain it to me, i.e. what am I doing wrong?
> Below I excerpted a little more from two newfs tries in case the
> additional info helps someone sort out what is going wrong.
> Any clues or pointers to reference materials that actually explain
> what is going on are greatly appreciated.
> thanks,
> Adam


This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:13:21 CDT