SUMMARY: PANIC (cpu 0): bs_frag_alloc: pingp failed

From: lombardi emanuele <lele_at_mantegna.casaccia.enea.it>
Date: Fri, 06 Nov 1998 12:20:02 +0100 (MET)

Dear friends,

Thanks to this marvellous mailing list I've solved the problem reported
after my signature.

The problem was probably due to an AdvFs bug which is said to be solved by the
patch #2 of Dunix 4.0d (duv40das00002-19980717.tar)
It seems that domains created before the #2 patch was applied
MAY run into the above problem.

The solution is simple to say , but heavy to do:

1) apply the #2 patch
2) backup all your data
3) umount each fset
4) delete each domain
5) (otional ?) a) scu format the SCSI disks
                b) re-initialize Logical Drive for RAID SWXCR disks
6) re-create each domain and fset
7) re-mount each fset
8) restore all your data

It worked for me (up to now, at least: finger crossed!)
To be sure to have a clean situation I did the above point 5) but I'm
not sure it was necessary.
Since my root & usr filesystems are both AdvFs, I had to use a
secondary boot disk (which I already had) to recreate the normal root &
usr domains in the above way.

Special thanks to whom replyed:

        Joanna Gaski <jgaski_at_wpi.edu>
        alan_at_nabeth.cxo.dec.com
        Charles M. Richmond" <cmr_at_iisc.com>
        rajeshar_at_swecexc.xko.dec.com
        Knut Hellebø <Knut.Hellebo_at_nho.hydro.com>
        soma_c_at_decus.fr (Claude SOMA - CNTS)

Long life to our mailing list,
Greetings from Italy,

Emanuele

-- 
 Emanuele Lombardi
 mail:  AMB-GEM-CLIM ENEA Casaccia
        I-00060 S.M. di Galeria (RM)  ITALY
 mailto:lele_at_mantegna.casaccia.enea.it
 tel	+39 6 30483366 fax	+39 6 30483591
     This transmission was made possible by 100% recycled electrons.
On  4 Nov, lombardi emanuele wrote:
> Dear axp masters,
> 
> My AS 1000 running Dunix 4.0d (patch 1) 
> has suddently started crashing every 10 minutes more or less!!
> 
> The message on the console is the following:
> 
> 	PANIC (cpu 0): bs_frag_alloc: pingp failed
> 	N1=-1035
> 	synancing disks.... malloc failed: bucket size=1152 #of failures=1, ra0fffff  ...............
> 	455 malloc failed
> 	465 done
> 
> then the reboot is done and after a few minutes it happens again!!
> 
> All my fs are advfs and RAM is 128Mb (I know is few, but it worked for
> the last 3 years!!)
> 
> I went to Single User mode and stay there for a couple of hours and NO
> crash happened, But going back to multiuser (switchin the server off to
> be sure) did'nt change the situation: it crashes again!
> 
> While in Single User I verify-ed all FS and had no problem
> and I defragment all fs and had no problem again.
> 
> 
> Can you, please, help me ?
> 
> Thanks and greetings from Italy,
> 
> Emanuele
> 
> ---
> 

Received on Fri Nov 06 1998 - 11:23:08 NZDT