Relax-and-Recover Frequently Asked Questions
Upgrading Support
Question: rear upgrade fails via rpm -U
If you ever have this behavior you better first remove the old version with rpm -e and then do a fresh installation again with rpm -ivh or yum install rear
Hardware support
mkrescue support
Question: Not enough disk space available in /boot for GRUB rescue image?
If you see above error message then you can edit the /etc/rear/local.conf and add
GRUB_RESCUE=
to avoid the rear generates a rescue image under the /boot/grub directory. By default,
rear is not anymore generating a rescue image and adding it to your grub configuration.
General Questions
Question: cannot make pipe for command substitution: Too many open files
Your linux system linits the amount of open files for the user root to a default which is too low (usually 1024). You should increase it - see details in issue #586
Question: How to create a bootable USB disk from an ISO image
Using the BACKUP=NETFS and OUTPUT=ISO method prodices an ISO image to boot from in rescue/rcover mode. However, suppose the system to recover (or clone) cannot use an ISO image to boot from, but does have an USB interface which could be used to boot from. The easiest way would be to copy the ISO image to an USB stick and boot from the USB stick, right?
How can we do this? You can use UNetbootin to accomplish this. Only a few steps are required:
sfdisk /dev/sdd <<EOF
;
EOF
mkfs.vfat -F32 -n REAR-USB /dev/sdd1
Pull and replug the USB stick. Verify it is mounted.
unetbootin method=diskimage isofile=/path/to/[ISO].iso installtype=USB targetdrive=/dev/sdd1 autoinstall=yes
More information about unetbootin can be found at http://sourceforge.net/p/unetbootin/wiki/commands/
Question: Cannot create ISO image larger then 4 Gb
Indeed that is try by default setting within rear. We prefer to have smaller boot image on ISO. However, when combining the tar archive and rescue image on the same ISO then it could happen that 4 Gb is not enough. To solve this read issue #581
Backup support
System migrations
IP Migration
Question: Can we define a fixed IP address with a rescue image?
Yes, you can do the following on the production system (before you run rear mkbackup). Create a file /etc/rear/mappings/ip_addresses with the following content:
eth0 192.268.1.100/24
Of course, replace the above IP address and cidr with your settings. You can also define eth0 dhcp to force DHCP when you boot with the rescue image.
There a second file /etc/rear/mappings/routes with the following content:
default 192.168.1.1 eth0
As you can guess this is used to define the default gateway (also here replace the items with your settings).
And, if you forgot these files you can still define a fixed IP address when you boot up from the rescue image. You need to interrupt the automatic boot process and give some extra kernel options:
ip=192.268.1.100 nm=255.255.255.0 gw=192.168.1.1 netdev=eth0
Question: Can we force a static IP address with the rescue image?
Yes, even when you are currently using DHCP you can define the variable
USE_STATIC_NETWORKING=1
in the configuration file /etc/rear/local.conf. You might also consider to define a static IP address in /etc/rear/mappings/ip_addresses (see previous question).