| Age | Commit message (Collapse) | Author |
|---|
|
The new classes 'DataSourceNoCloud' and 'DataSourceNoCloudNet'
implement a way to get data from the filesystem, or (very minimal)
data from the kernel command line. This allows the user to seed data to
these sources.
There are now 2 "cloud-init" jobs, cloud-init-local that runs on
mounted MOUNTPOINT=/
and 'cloud-init' that runs on
start on (mounted MOUNTPOINT=/ and net-device-up IFACE=eth0 and
stopped cloud-init-local )
The idea is that cloud-init-local can actually function without network.
The last thing in this commit is "uncloud-init".
This tool can be invoked as 'init=/usr/lib/cloud-init/uncloud-init'
It will "uncloudify" things in the image, generally making it easier
to use for a simpler environment, and then it will exec /sbin/init.
|
|
device names presented in the metadata service may not be what the kernel
has named them. This can be for more than 1 reason. But some examples:
- device is virtio, metadata named 'sd'
- device is xvdX, metadata named sd
Those are the two situations that are covered here. More complex, but
not covered are possibly:
- metadata service named device 'sda1', but it should actually be 'vdb1'
LP: #611137
|
|
|
|
Previously, all you would get was a warning to the console on config
module failure. This changes to get a stack trace of the failure to the
console, which is much easier for debugging.
|
|
|
|
|
|
ec2-run-instances
--block-device-mapping /dev/sdd=:1
--block-device-mapping /dev/sde=snap-4cda7b24
--block-device-mapping sdf=snap-d4d90bbc
resulted in:
'block-device-mapping': {'ami': '/dev/sda1',
'ebs1': '/dev/sdd',
'ebs2': '/dev/sde',
'ebs3': 'sdf',
'ephemeral0': '/dev/sda2',
'root': '/dev/sda1',
'swap': 'sda3'}
|
|
|
|
It just seems like for cloud instances, getting /etc/fstab written
incorrectly with the result of non-booting system is worth avoiding.
|
|
What caused this was having an open ended list on what "names" might be
found in the metadata service. That list is now trimmed down to the
the following values:
ephemeral*
root
ami
swap
The above list was found from crawled medata data services in the latest
maverick tests I did. The following is the complete list of entries that
were there:
'ami': '/dev/sda1',
'ami': 'sda1',
'ephemeral0': '/dev/sda2'
'ephemeral0': '/dev/sdb'
'ephemeral0': 'sda2'
'ephemeral0': 'sdb'
'ephemeral1': 'sdc'
'ephemeral2': 'sdd'
'ephemeral3': 'sde'
'root': '/dev/sda1'
'root': '/dev/sda1'
'swap': 'sda3'
Also, this limits which devices will have "/dev/" prepended to them to
sda, sda1, xvda, xvda1, hda1, hda, vda.
LP: #603329
|
|
|
|
The starts-with determination of mime type was overriding an explicit
setting in the mime-type. This was evident when the mime type specified
boothook, but the content began with '#!'. In that case, the content
would run as a user script rather than boothook.
LP: #600799
|
|
|
|
The goal was to remove '#cloud-boothook' from a part if the part
started that way. This was to allow user data of
#cloud-boothook
#!/usr/bin/perl
...
to be handled correctly. That had 2 bugs
1.) the prefix string was wrong
2.) was checking for '\r' in the wrong location
|
|
If user gives bad cloud-config syntax, its not very useful to die, as
that is most likely to leave the system unreachable. This instead
logs the error and continues as if it no cloud-config was given.
|
|
|
|
|
|
|
|
|
|
LP: #582667
|
|
|
|
to be debug (with traceback). The exception is still raised, but
no reason for the whole traceback to be on error
|
|
This is useful for getting a config option that is either string or a
list as a list
|
|
568139 was fixed because the test for "always" was using "is"
instead of "=="
LP: #568139
|
|
|
|
Previously, most of the config semaphores were prefixed with 'config-'.
Ie, a sem/ list would look like:
apt-update-upgrade.i-7c908817
config-misc.i-7c908817
config-mounts.i-7c908817
config-puppet.i-7c908817
config-ssh.i-7c908817
consume_userdata.i-7c908817
disable-ec2-metadata.always
set_defaults.i-7c908817
set_hostname.i-7c908817
With the last release (0.5.11), those config- would have been removed.
I'll handle this correctly yuckyness in the ubuntu package upgrade
(avoiding re-running scripts that were already ran)
|
|
|
|
passing the instance-id of this instance to a boothook will give it
the unique id that is needed to implement run-once-per-instance.
|
|
if user data is of type text/cloud-boothook, or begins with
#cloud-boothook, then assume it to be code to be executed.
Boothooks are a very simple format. Basically, its a one line header
('#cloud-config\n') and then executable payload.
The executable payload is written to a file, then that file is executed
at the time it is read. The file is left in
/var/lib/cloud/data/boothooks
There is no "first-time-only" protection. If running only once is
desired, the boothook must handle that itself.
|
|
The list of cloud-config modules is now kept in cloud config itself.
There is a builtin list in cloudinit, which is overrideable by
/etc/cloud/cloud.cfg or user data cloud-config.
This should make the modules more easily added or removed (as no code
needs to be edited now)
Basic summary of changes:
- move CloudConfig.py -> cloudinit/CloudConfig/__init__.py
- split cloud-config modules into their own files named
cloudinit/CloudConfig/cc_<name>.py
- remove all the upstart/cloud-config-* scripts, replacing them with
upstart/cloud-config.conf
|
|
use get_base_cfg from CloudInit:read_cfg
|
|
|
|
This logging infrastructure in cloudinit:
- uses python logging
- allows user supplied config of logging.config.fileConfig format to be
supplied in /etc/cloud/cloud.cfg or in cloud_config by user data.
- by default, tries to use syslog, if that is not available, writes directly to
/var/log/cloud-init.log (syslog will not be available yet when cloud-init
runs)
- when using syslog, the doc/21-cloudinit.conf file provides a rsyslogd
file to be placed in /etc/rsyslog.d/ that will file [CLOUDINIT] messages
to /var/log/cloud-init.log
|
|
|
|
Also
- adds some debugging information when its waiting
- add 'uptime' printout on initial cloud-init invocation
|
|
LP: 507709
LP: #507709
|
|
|
|
On first boot of an instance, cloud-config replaces
/etc/apt/sources.list with references to a local mirror. This will also
rename the old list files in /var/lib/apt/lists .
LP: #513060
|
|
This marks the comment option of fs_mntops in fstab (man fstab) with
cloudconfig for each of the mount options added by cloudconfig. It will
search through existing lines, any entry written by cloudconfig will
be deleted.
LP: #524562
|
|
In instance store I was used to block-device-mapping:
{'ami': 'sda1',
'ephemeral0': 'sda2',
'root': '/dev/sda1',
'swap': 'sda3'},
in ebs, image registered with
'--block-device-mapping /dev/sda2=ephemeral0',
metadata is showing block-device-mapping:
{'ami': '/dev/sda1',
'ephemeral0': '/dev/sda2',
'root': '/dev/sda1',
'swap': 'sda3'},
Without this change, 'ephemeral0' would not get translated and would
show up in /etc/fstab.
|
|
would not be converted to 'ephemeral0'
|
|
|
|
The cache file location is not instance specific. As such, if it is
not cleaned from the image, a re-bundle would get the old data. To avoid
that, clear the cache in cloud-init.
|
|
|
|
on object cache stored there.
|
|
|
|
|
|
runcmd allows simple running of commands at rc.local like time frame
see doc/examples/cloud-config.txt for more info.
|
|
LP: #523832
|
|
|
