Technical article 126407

One or more of the DDB files are not present/corrupt/inaccessible.

OpenServer 6.0.0

Written · Last revised

Problem

At boot time after the HBA configuration is displayed on the console I get the message:

One or more of the DDB files are not present/corrupt/inaccessible.

Solution

Your system maintains backup copies of the DDB files. Before attempting to create new device tables, check if the backup copies of the device tables are still sane byt doing the following as root:

cp /etc/Odevice.tab /etc/device.tab
cp /etc/security/ddb/Oddb_dsfmap /etc/security/ddb/ddb_dsfmap

Now try running a command that uses the DBB files like:

devattr cdrom1

If this command succeeds then you have cleared the problem. If the above procedure did not correct the problem, execute the following shell script as root:

#!/sbin/sh
rm -f /etc/device.tab /etc/security/ddb/ddb_dsfmap 2>/dev/null
if [ $? -ne 0 ]
then
        echo "Could not remove device tables"
        exit 1
fi
echo "MAGIC%NO 1" 2>/dev/null | tee /etc/device.tab > \
     /etc/security/ddb/ddb_dsfmap 2>/dev/null
if [ -x /etc/scsi/pdimkdev -a -x /etc/scsi/pdimkdtab ]
then
        /etc/scsi/pdimkdev -fi 2>/dev/null && \
           /etc/scsi/pdimkdtab -fi 2>/dev/null
        if [ $? -ne 0 ]
        then
                echo "Could not create disk entries in device tables"
                exit 1
        fi
else
        echo "Could not run pdi commands"
        exit 1
fi
if [ -x /usr/sadm/sysadm/bin/mkdtab ]
then
        /usr/sadm/sysadm/bin/mkdtab 2>/dev/null
        if [ $? -ne 0 ]
        then
                echo "Could not make floppy entries into device table"
                exit 1
        fi
else
        echo "Could not execute mkdtab"
        exit 1
fi
echo "Device table has been rebuilt"

If you received this article via fax and don't want to type the entire script above, execute the following commands:

(Don't type # -- this is to indicate your shell prompt )

# rm -f /etc/device.tab /etc/security/ddb/ddb_dsfmap 2>/dev/null
# echo "MAGIC%NO 1" >/etc/security/ddb/ddb_dsfmap
# echo "MAGIC%NO 1" > /etc/device.tab
# /etc/scsi/pdimkdev -fi
# /etc/scsi/pdimkdtab -fi
# /usr/sadm/sysadm/bin/mkdtab

Ensure that the HBA for your root disk is also available in the kernel.

For example, if the root disk is stored on an "ida" HBA then check it is in the kernel with:

# strings /stand/unix | grep ida
If not match is found for the "ida" driver then it can be statically
linked in by editing the file /etc/conf/sdevice.d/ida and adding a
$static line as:
$version 2
$static
ida ....

The kernel should then be re-linked with:

# /etc/conf/bin/idbuild -F
# init 6
NOTE:    If /var has not been mounted then using "vi" will result in a failure
         looking for /var/preserve.
To get around this problem run:
# mkdir /var/preserve
# chmod 1777 /var/preserve
# chown bin:bin /var/preserve
Then "vi" will work.
NOTES:   The /etc/device.tab and /etc/security/ddb/ddb_dfsmap files are also
         updated with protocol entries by the post install script of the inet
         package (eg. arp, icmp, igmp, incf, ip, ipip and llcloop) and with
         pseudo device node entries by the post install script of the acp
         package (and via netcfg).  These entries are not recreated using the
         commands in this TA (and so you may notice that the new device
         databases are smaller than the originals).
These additional entries are only used for enhanced security features
and their absence does not cause any problems with the normal running
of the system.  If they are required to be in the device database then
a non-corrupt device database should be recreated as above and then
the relevant packages removed and reinstalled (including any relevant
updates.  Alternatively, a non-corrupt device database can be restored
from a recent system backup.

From the Xinuos technical article database, record 126407. Reproduced with permission. The original is at www.sco.com/ta/126407, which is reachable but blocked from search engines.

Next step

Need this fixed rather than explained?

Tell us the release, what the system runs and what it is doing. We will reply by email with what we would do next.

  • You deal with the people who do the work
  • No obligation, and no upgrade required before we will help
  • We tell you plainly if something is not worth doing

Send an enquiry

We reply by email. If a server is down, say so in the first line. You can also write to help@scoexpert.com.