OpenBSD Manual Partitioning at Installation

When installing OpenBSD, at the partitioning section, the installer asks which option to choose:

1. Use (A)uto layout
2. (E)dit auto layout
3. Create (C)ustom layout

The first "a" is a good choice to rely on the default settings and save operation time at installation.

I sometimes, however, like to choose manual partitioning.
For example, using storage-limited virtual machines, I want to let the sizes of the specific mount points such as /var and /home as big as possible.
At such a time, "e" or "c" is also a good choice.

• "e": Use auto layout but configure it.
• "c": Create my own layout from the beginning.

The choices are followed by disklabel running:

We can manage partitions via these commands:

? | h                # show help
2.7G
p                    # print partitions

a [Partition Label]  # add partition
d [Partition Label]  # delete partition
z                    # delete all partitions

x                    # exit & lose changes
q                    # quit & save changes

Today, I created OpenBSD vm image for Google Compute Engine (GCE) whose storage size is limited to 30 GBytes.
I changed mount points and their sizes like these operations:

> a a                  # add parition labeled as 'a'
offset: [64]           # just click Enter key
size: [62910476] 1.0G  # set partition size
FS type: [4.2BSD]      # just click Enter key
mount point: [none] /  # set mount point

> a b                  # add parition labeled as 'b'
offset: [2104512]      # just click Enter key
size: [60806028] 1.1G
FS type: [swap]        # just click Enter key

# skip 'c'

> a d                  # add parition labeled as 'd'

...

> p                    # check partitions in the end
> q                    # save changes and exit

Finally, I changed the whole disk like this:

Thus, I built OpenBSD on GCE and launched a small database server with its storage big enough for my usage : )

Updated:

These assignments are also good for a tiny virtual machine when serving with using Google Cloud SQL as database : )