| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This adds a DataSource for DigitalOcean's metadata service. The service is
documented at https://developers.digitalocean.com/metadata/ .
|
|
|
|
|
|
Without this change the tests are currently failing on rhel7
since a location where a hostname file is written no longer
exists at that location when systemd is active. To avoid this
allow the test to inspect if the distro has systemd enabled
and avoid testing the file when systemd is being used so the
test passes.
We likely need to figure out a better way to test features
that no longer exist as files but exist as commands with
systemd in general.
|
|
|
|
|
|
There are a couple new test cases that are now erroring out.
- A usage of the non helper unit testcase base which is trigging
and error on python 2.6 due to lack of method, fix this by using
the base helper class.
- A freebsd distro test check thats looking for /etc/resolv.conf and
examining its contents, which won't exist due to our mocking routine
that does not allow that file to be read.
- A freebsd distro test where the distro class tries to call into
['ifconfig', '-a'] for a set of values, those values don't exist on
the machine I am running on (and likely others machines) so we should
mock the subp function out (that the distro class calls) and correctly
return values that will work for the testcase.
|
|
|
|
For now, this vendor data handling is just added to openstack.
However, in an effort to allow sanely handling of multi-part vendor-data
that is namespaced, we add openstack.convert_vendordata_json .
That basically takes whatever was loaded from vendordata and takes
the 'cloud-init' key if it is a dict. This way the author can
namespace cloud-init, basically telling it to ignore everything else.
|
|
instead of taking a version that they should look for,
the readers now just select the highest supported version.
definitely a use case later for having version= but nothing
is using it now.
|
|
make pyflakes now passes.
|
|
|
|
|
|
|
|
|
|
This just removes comments '# pylint:' things and other code
remnents of pylint.
|
|
Fixed all complaints from running "make pep8". Also version locked
pep8 in test-requirements.txt to ensure that pep8 requirements don't
change without an explicit commit.
|
|
|
|
This change works around a bug in CloudStack's EC2 api implementation.
That is filed upstream at [1].
The fix is safe for openstack and EC2 use cases as well.
EC2 and OpenStacks' EC2 metadata service both return a list with
access to either of:
<url_base>/latest/meta-data
<url_base>/latest/meta-data/
Additionally, the responses explicitly contain a trailing '/' for
items that have a child. The ec2_utils code then just re-uses the trailng
/ there. Thus, only the top level request for 'meta-data/' needs
the explicit fix.
This also changes test cases. Those test cases failed without fixing
them. If ever this regressed, those would fail again.
--
[1] https://issues.apache.org/jira/browse/CLOUDSTACK-7405
LP: #1356855
|
|
|
|
previous commit occurred because the selinux test was failing
in a schroot where there was no /etc/hosts.
Now, fix that test more correctly, and fix some bad assumptions in
the SeLinuxGuard.
|
|
Also, in debian packaging depend on it (so it wont skip there).
|
|
|
|
|
|
previously this would fail:
http_proxy=http://foo.bar make test
now it will pass. This works around a bug where httpretty is not able to
patch http operations if http_proxy is set.
https://github.com/gabrielfalcao/HTTPretty/issues/122
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
A few of the current tests have been continually failing
in python 2.6 based systems, due to lack of unit test
functions that are now added to ensure we can run the
unit tests (and not have to ignore those failures) on
python 2.6
|
|
|
|
LP: #1316597
|
|
On systems with a ttyS1 and nothing attached, the read attempts
that the cloud sigma datasource would do would block.
Also, Add timeouts for reading/writting from/to the serial console
LP: #1316475
|
|
LP: #1303986
|
|
|
|
|
|
|
|
Instead of just trying to see if userdata decodes as the indication that
it should be encoded, the user must explicitly set this.
The "just try it" will fail in the case where the user had other use
of user-data and wanted a blob of data to go through unrecognized by
cloud-init.
In cases where there can be mistake in automatic behavior,
and some users may be relaying on old behavior, its best to just require
explicit use.
|
|
|
|
|
|
|
|
This adds the ability to read a 'base64_fields' entry in the metadata,
and if cloud-init-userdata is listed in that, then content will be
base64 decoded first.
|
|
This extends 'random_seed' top level entry to include a 'command'
entry, that has the opportunity to then seed the random number generator.
Example config:
#cloud-config
random_seed:
command: ['dd', 'if=/dev/zero', 'of=/dev/random', 'bs=1M', 'count=10']
LP: #1286316
|
|
there was a bug that prevented seeding of /dev/urandom from metadata provided
by the datasource unless the user provided random_seed config.
This should, instead, be the default behavior.
|
