On Wed, Oct 08, 2003 at 09:27:18AM -0400, Dennis McRitchie wrote:
> Hi Ralf,
> 
> I'm sure you don't want to hear this given how much time you spent getting rid of 
> this stupid error :-), but it doesn't
> look like hanging rpm processes are causing the problem. I ran a ps -ef on both 
> Solaris 9 systems, and they are clean.
> Here is the exact output from --db-rebuild and --db-cleanup.
[...]
> I ran truss -f on it and, as expected, found references to EAGAIN (errno 11) from 
> mmap calls in it, presumably generated
> by rpmdb since truss without the -f did not show any errors. I can send you the 
> truss output if you think that would
> help.
> 
> I also tried using Jeff Johnson's suggested solution of adding the following line to 
> rpmmacros, and that fixed the
> problem as expected, but it's not the ideal way to go, of course:
> 
> %__dbi_cdb create cdb mpool mp_mmapsize=16Mb mp_size=1Mb private
> 
> Any thoughts about what I should try next?

There was a similiar problem in the past. $prefix/RPM/DB was on a
NFS volume. What filesystem type is used in your case.


   (mk)

-- 
Matthias Kurz; Fuldastr. 3; D-28199 Bremen; VOICE +49 421 53 600 47
   >> Im pr�motorischen Cortex kann jeder ein Held sein. (bdw) <<
______________________________________________________________________
The OpenPKG Project                                    www.openpkg.org
User Communication List                      [EMAIL PROTECTED]

Reply via email to