| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If btrfs is not present in the kernel we end up with strange output like
this:
Error creating disk image2014-06-10 16:00:40
[devel-system-x86_64-generic][my-raw-disk-image][rawdisk.write]Failure
to create disk image at /src/tmp/testdev.img
ERROR: Command failed: mount -o loop /src/tmp/testdev.img
/src/tmp/deployments/tmpQ7wXO1/tmp4lVDcu/tmpvHSzDE
mount: mounting /dev/loop0 on
/src/tmp/deployments/tmpQ7wXO1/tmp4lVDcu/tmpvHSzDE failed: Device or
resource busy
To avoid this confusing error, Morph should explicitly check first.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If INITRAMFS_PATH is specified and the file exists, then the produced
kernel command line will use root=UUID=$uuid_of_created_disk rather than
root=/dev/sda, which may be incorrect.
Help files have been updated to mention the new option.
This leads to an unfortunate duplication of the path to the initramfs,
in both the location field of the nested deployment and the
INITRAMFS_PATH of the disk image creation.
However, an initramfs could be produced by a chunk and put in the same
place, so it doesn't make sense to couple the rawdisk and initramfs
write extensions to remove this duplication.
Similarly, there may be multiple valid initramfs in the rootfs e.g.
extlinux loads a hypervisor, which is Linux + initramfs, and the
initramfs then boots a guest Linux system, which uses a different
initramfs.
This makes it important to explicitly let the rootfs write extensions
know which to use, or not as the case may be.
util-linux's blkid is required, since the busybox version ignores the
options to filter its output, and parsing the output is undesirable.
Because syslinux's btrfs subvolume support is limited to being able to
use a non-0 default subvolume, the initramfs has to be copied out of
the run-time rootfs subvolume and into the boot subvolume.
This pushed the required disk space of a minimal system over the 512M
threshold because we do not have the userland tooling support to be able
to do a btrfs file contents clone.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This creates a gzipped cpio archive that may be used as an initramfs.
It is hard-coded to use gzip to compress the initramfs, since it's the
most common way to do it.
This is unfortunate, since the busybox gzip utility only allows maximum
compression, which is rather slow and doesn't give progress reporting,
so you can easily think it's gotten stuck.
It's possible to use other compression formats, but they need the kernel
to be built with them supported, and in the case of lz4, unusual userland
tools to create it, since the version of lz4 supported in the kernel is
not what the standard lz4 tools produce.
|
|\
| |
| |
| |
| | |
Reviewed-By: Richard Maw <richard.maw@codethink.co.uk>
Reviewed-By: Lars Wirzenius <lars.wirzenius@codethink.co.uk>
|
| | |
|
| | |
|
|/
|
|
|
|
|
| |
Add support to the VirtualBox write extension to notice if we are
doing a Vagrant Basebox installation and not do the clever network
setup we normally do to allow machines to talk to one another since
this confuses Vagrant quite a bit if it is left in.
|
|
|
|
|
| |
Move the help out of the comment and into a help file,
and add a clearer example.
|
| |
|
|
|
|
|
| |
Test that a VM with the given name does not already exist, and check
that the files specified in ATTACH_DISKS do already exist.
|
|
|
|
|
| |
Also, change it to log the real error message in morph.log before
raising a more general exception to the user.
|
|
|
|
|
|
| |
Slight duplication is necessary, but it's only a few lines. We could
move the duplicated code into the base class in 'morphlib.writeexts' if
there was more duplication.
|
|
|
|
|
|
| |
Begin adding help documentation for configuration
and write extensions starting with nfsboot.write,
rawdisk.write and tar.write.
|
| |
|
|
|
|
|
| |
We were building it from a variable in some places and hardcoding it in
others; now we build it from a variable everywhere.
|
|
|
|
|
|
|
|
|
|
|
| |
Move some code to the '.check' extension to verify that the deployment
can happen *before* spending 5 minutes unpacking and configuring the
rootfs.
This is not a perfect solution yet because when multiple systems are
being deployed in a cluster, we do not check all systems and then
deploy them all. Instead, we check one, then deploy it, then check the
second, etc.
|
| |
|
|
|
|
|
|
|
|
|
| |
There is no longer a default /etc/fstab in the Baserock fhs-dirs chunk,
and the nfsboot.write extension does not use the default Btrfs system
layout so no entry is added for / to /etc/fstab at deploy-time.
We cannot have / in /etc/fstab for nfsboot deployments because it
causes systemd to remount / during bootup, which breaks everything.
|
|
|
|
|
|
|
|
|
|
|
|
| |
This moves the deployed system to somewhere on the host.
Any existing contents of the directory is deleted, so don't try to be
clever and deploy a new system on top of / in place of a proper upgrade.
It can be used to deploy a chroot, sysroot or container, but its
current use is to allow for nested deployments to include another
system in itself, since the parent deployment's "$1" is prepended to
the sub-deployment's "$2".
|
|
|
|
|
|
|
| |
If the disk image was not yet created then the os.remove() call fails
and the original exception gets lost, causing confusion and sadness.
Also print status earlier on failure
|
|
|
|
|
| |
Most write extensions don't handle both initial deployments and upgrades
of a system.
|
|
|
|
|
| |
Now you can deploy an upgrade, set it to be the default version and reboot
into it all with one call to `morph deploy`.
|
|
|
|
|
| |
Also, be more flexible when parsing environment booleans -- convert to
lower case and match 0/1 and true/false as well as yes/no.
|
| |
|
|
|
|
|
|
| |
We now have a OS version manager tool in Baserock (in tbdiff.git). The
code to deploy a new base OS version should live there, to minimise
duplication between write extensions.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The shared state directories defined in writeexts.py (/var, /home etc.)
are now separate Btrfs subvolumes that are mounted in place using fstab.
There are some warnings on mounting /var and /srv about the mountpoint
not being empty. Not yet investigated.
If a configure extension has already added / to the fstab, use the
device it chose rather than assuming /dev/sda. This is required for the
vdaboot.configure extension that we use for OpenStack deployments.
Similarly, if a configure extension has added an entry for a state
directory in /etc/fstab already, we don't replace it with a /state/xxx
directory. That's only done as a default behaviour.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
VirtualBox changed a command line option in 4.3 incompatibly, so we now
have to check the version number and change an option from --sataportcount
to --portcount if the version of VirtualBox running on the target is at
least 4.3
This turns the version into a tuple and compares it against another,
since it's more reliable than comparing strings, which will count '1.10'
as earlier than '1.2', and more convenient than comparing the digits
individually.
|
|
|
|
|
|
|
|
|
|
|
| |
This option lets the install-files config extension overwrite existing files.
A file will only be overwritten if the overwrite flag is specified for that file.
Since the overwrite arg is optionally prepended to the manifest line,
this patch should not break existing manifests
With this patch default config files can be replaced with project specific
config files
|
|\ |
|
| | |
|
|\ \
| | |
| | |
| | |
| | |
| | |
| | | |
Reviewed-by: Richard Maw
At his suggestion, fixed the call to sorted() to be a call
to asciibetical().
|
|/ /
| |
| |
| |
| | |
This will allow the user to append text to /etc/fstab during a
deployment, without having to write custom configuration extensions.
|
|/ |
|
|
|
|
|
| |
This snuck in since the test suite could not be run when TMPDIR was
on a tmpfs.
|
|
|
|
|
|
|
|
|
| |
openstackssh.write: Write extension which deploy a raw image of baserock
directly to an OpenStack machine using python-glanceclient. The raw image
deployed has modified its bootloader to use virtio disks.
vdaboot.configure: Configuration extension to change the mount point of
"/" to use virtio disks (/dev/vda).
|
|
|
|
| |
Patch from Paul Sherwood.
|
| |
|
| |
|
|
|
|
| |
systems.
|
|
|
|
|
|
|
|
|
| |
Accordingly the rsync manual:
"A trailing slash on the source changes this behavior to avoid
creating an additional directory level at the destination. You can
think of a trailing / on a source as meaning "copy the contents of
this directory" as opposed to "copy the directory by name".
|
|
|
|
| |
Unmounting the root disk as the side effect of turn it to be read only
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Or else this
ln -s -f update1 /mp/systems/default
will do this
'/pp/systems/default/update1' -> 'update1
When we want
'/pp/systems/default' -> 'update1
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
-s, or --protect-args prevents the file path components of destination
or source paths being interpreted by the remote shell.
This is for wildcards or other shell features, but it breaks when paths
have whitespace.
We tend to always use absolute paths, so all uses of rsync now pass -s.
kvm.write needs it, since the disk can be written to a path with spaces.
Nfsboot and ssh-rsync need it because version labels are used, which may
have spaces, and temporary directories are used, which could have spaces
in weird TMPDIR configurations.
|
| |
|
| |
|
|
|
|
|
|
|
| |
Verify if are using and older extlinux configuration and upgrade it
if the case, by checking if the "default" symbolic link exists on the target.
Note that with the symbolic link we don't need to update extlinux configuration after an upgrade
|
|
|
|
| |
systems.
|
| |
|