Booting diskless/diskful/mixed systems with an initramfs

Although not required, using an initramfs to bootstrap oneSIS nodes can provide more flexibility than traditional NFSroot methods. The mk-initramfs-oneSIS script (described in section 6.4) is capable of automatically building an initramfs that can be customized for any kind of node, or generalized for an entire cluster of nodes.

Once the initramfs is created, nodes can boot into diskless/diskful/mixed environments by specifying the initramfs on the kernel command line:

• initrd=initramfs-3.10.0

The primary job of the initramfs is to bring up a network interface, configure the interface via DHCP, set the node's hostname, and mount its root filesystem. Specific kernel modules can be loaded by supplying options to mk-initramfs-oneSIS.

The root filesystem itself can be mounted via NFS or from a local disk. When a portion of the root filesystem has been deployed on a local disk, the initramfs can be configured to automatically mount those partitions before pivoting into the root filesystem.

An entire cluster (or any individual node or group within it) can be configured any way you want. The default initramfs template can also be extended to support almost any conceivable creative boot method.

The standard mkinitramfs or mkinitrd utility supplied with the distribution can still be used to bootstrap diskful nodes in many scenarios.

Note: Some distributions may not provide a kernel with CONFIG_PACKET support built directly into the kernel. If this is the case you will need to make sure to include the af_packet module when building an initramfs in order for the built-in DHCP client to work.