mirage-parser.c: In function 'mirage_parser_load_image':

If you try to compile libmirage then you might get the following error:

Obtaining CPU information in AIX

To obtain CPU information use the AIX command prtconf: 

WORM media with mhvtl and bacula

WORM (write once read multiple) media is necessary if you need to make sure that special backups are kept forever (or as long as possible). How long a backup must kept alive depends on the data to backup, company needs, government regulations and so on. To play a little with WORM media you can use mhvtl (I used the web gui to create a new media). If you are lazy then you can spare this step, you can use any other media as WORM. The advantage of a real WORM media is that it won'toverwrite in other libraries, environments or even accidently. A cheap example with less capacity for a WORM media are writable CD's and DVD's - not rewritable!

umount: /mnt busy

When trying to unmount a file system you may got the typical error like:

# umount /mnt
umount: /mnt busy

DNS lookup in AIX

To configure AIX for use with DNS create the /etc/resolv.conf first:

Configuring NTP in NetBSD

To configure the NTP client in NetBSD edit ntp.conf file end jump to the end of the file. Then uncomment all servers and add your own NTP server like this:

Obtaining CPU information in HP-UX

To obtain CPU information use the HP-UX command machinfo: 

Migration with bacula

Imagine you have the following situation: two database servers each with a large database. To backup each database you need 12 hours but you only have one tape drive. How to backup both databases to one tape drive without running the backup at daytime? As the topic says it: using bacula migration. When using migration you can start both backups at eg. 20:00, one backup writes directly to the single tape drive, the other backup to a disk storage. In this case both backups are accomplished between 20:00 and 08:00 the next day.

Sun StorEdge 6120 password reset

Got a few Sun StorEdge 6120 without any password (the shipped one was not right). To access the 6120 I have to reset the password first. Here is a way how I accomplished it. First shutdown the 6120 and disconnect all power cables. Then use a screw driver to open the screw that holds the U1 controller on the back of the 6120. Then pull the U1 controller out of the 6120. In the top right corner there are a few jumpers gathered to J10. Use a jumper to connect the pins 14 and 16 like the image:

Using bacula with mhvtl

In a prior article I showed you how to setup a virtual tape library with mhvtl. I configured a MSL600 with 48 tapes. This device I want to use with bacula now. My current setup is the bacula director running on a Solaris 10 X86 machine, but the mhvtl is running on a Slackware X86 machine. To continue I have to install the bacula storage daemon on the Slackware machine first. To install the storage daemon you have to build the complete bacula software. Download a copy of the source and store them under /usr/src. Then extract the source package and change into the new directory: 

LDAP backup with bacula

To backup your LDAP with bacula is very easy. You only need to create a script that exports the DIT in a file. Then create a job that backups this file. To start go into the bacula configuration directory and define a new job:

Using a Sun StorEdge D240 as bacula storage

Yesterday I got a Sun StorEdge D240. Completely with cables, SCSI card etc. The D240 has two 36GB harddisks, a DDS4 tape drive and a DVD-ROM. Perfect to use it in my bacula server.
After installing the SCSI card and connecting the D240 I booted my Solaris 10 X86 machine and installed all devices (I'm using the Split Bus configuration). To install the necassery devices run the devfsadm command and use the iostat command to check the devices: