| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|\ \
| |/
|/|
| | |
0.24.x
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
value and added comments that query will now always do so.
The query method is only called in two places, from inside 'install' in
yum.rb and from inside 'uninstall' in rpm.
This behavior only happens when you have a lower version than the one you
are 'ensuring'.
Since the right package actually gets installed, the system is in sync
next time, but the event of installation is lost and subscribing resources
will never get it.
|
|\ \
| | |
| | |
| | | |
0.24.x
|
| | |
| | |
| | |
| | |
| | |
| | | |
Removed logic to reuse parser and log on server when there is a parsing error.
Now we just make a new parser and if there is an error, raise it up, which will result in errors on the client logs.
|
| |/
|/|
| |
| |
| |
| |
| | |
The tests were failing when run on a machine with certs on the file system
Stub out failure to read where appropriate
Worked fine at my desk :(
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
There was a save_object method in the http/handler module, but it
seemed unnecessary and was undocumented. I don't know I missed
it causing failures, but I've added it back with appropriate
documention.
|
| | |
|
| | |
|
| | |
|
|/ |
|
|\ |
|
| |
| |
| |
| |
| |
| | |
I wasn't testing whether the options where still a hash,
and an 'inject' loop wasn't returning correctly, so
the options weren't being set correctly.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
My request authorization code changed the request to set its 'node'
instance variable, rather than using its options, and I apparently
didn't change the Indirection class to do this.
I have *no* idea how these tests were passing before -- they passed for
me yesterday but failed for Andrew, and today they started failing for
me. Frightening.
|
| |
| |
| |
| | |
FAIL
|
|/
|
|
|
|
| |
Add the class to handle certs, specs and modify puppetd
should fix 1190, 1199, 1200
|
|\
| |
| |
| | |
git://github.com/lak/puppet into 0.24.x
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
All requests should now have an ipaddress add to them, they
should always be marked authenticated or not, and they should
have the certificate name set as their 'node' if a certificate
is present.
They both use the same methods they use for xmlrpc,
although there's no common code, to facilitate deprecation
of xmlrpc.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This basically consists of the node name (i.e., the certificate name),
the IP, and whether the request is authenticated (which is determined
by whether it provided a valid certificate).
Now the two REST classes (mongrel and webrick) need to correctly
pass this information to the indirection calls they make.
|
| |
| |
| |
| | |
the ResourceTemplate class.
|
| |
| |
| |
| | |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|/
|
|
|
|
| |
(i.e., client-side templates). This would really only be used
for composite resources that pass the results of the template
on to generated resources.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
More correctly handle indentation of lines enclosed in [] when the
contents themselves contain [], such as arrays of object references
for requires.
Exclude newlines from more of the parsing regexes for more accurate
results.
Fontify references to objects even if they're not at the start of a
line.
|
|\ |
|
| | |
|
| | |
|
|/
|
|
|
| |
Exporting or collecting resources no longer raises an exception
when no storeconfigs is enabled, it just produces a warning.
|
|\ |
|
| |
| |
| |
| | |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|\ \ |
|
| |/ |
|
|/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Master handler previously provided the support for the :node_name
setting, and that functionality has now been moved into the Node
class. At the same time, the names to search through have been
changed somewhat: Previously, the certificate name and the
hostname were both used for searching, but now, the cert name
is always searched first (unless node_name == facter), but only
the Facter hostname, domain, and fqdn are used otherwise. We no
longer split the cert name, only the hostname/domain/fqdn.
In the general case, this provides no behaviour change, because
people's hostname is the same as their certname. This only
results in a change in behaviour if you specify a certificate
name that is a normal node name, and you want to look that node
up by something other than the full name in the certificate.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|\ |
|
| | |
|
| | |
|
| |
| |
| |
| | |
This reverts commit d54338f8921b4b2973dcde52576d7e9fa58148cb.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
master branch.
|
|\| |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
catalogs.
It largely worked previously, except when aliases were set,
which require catalogs. This now converts all parser resources
to trans_objects then to RAL resources, and sets the catalog
in the meantime. This allows aliases to work just
fine when converting directly from parser catalogs to RAL catalogs.
|
| | |
|
|/ |
|
|
|
|
|
|
| |
Previously, we prefetched then generated, which caused generated
resources that needed prefetching not to work. This just reorders
the calls, so generated resources now get prefetched.
|
|\
| |
| |
| |
| |
| | |
Conflicts:
CHANGELOG
|
| |
| |
| |
| |
| | |
Previously, they failed if you tried to get them to autogenerate
an id, because they assumed that a result would be returned.
|