| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
| |
Previously, modules were not using their environments
when looking up their paths, which meant that they
often found files in the wrong environment.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
| |
Comments and multi-line comments produces no token per-se during
lexing, so the lexer loops to find another token.
The issue was that we were not skipping whitespace after finding
such non-token.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
| |
Bellman)
|
|
|
|
| |
Signed-off-by: James Turnbull <james@lovedthanlost.net>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
| |
We were previously throwing exceptions.
This also ports all of the tests for variable lookup
over to rspec.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
| |
Adding the tags to the rails collect query can reduce performance
because there are 2 more tables to join with.
So we make sure to include tags in the query only when it is
necessary.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
|
|
|
| |
The various collector specs covering the rails query code were not
in fact covering anything.
This patch fixes the specs to integration tests that actually verify
the rails query building code works fine.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
|
| |
This is so that overriding the "tag" metaparameter ends-up in the
resource tags on the server.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
| |
|
|
|
|
|
|
|
|
| |
This function acts exactly as the 'include' function, but also
adds an ordering relation between the included class and the class
where the require function is.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
This won't affect most people, but it's a good
default to have for those who can support it.
Signed-off-by: Luke Kanies <luke@madstop.com>
Minor changes to previous commit
|
|
|
|
|
|
| |
This provides about a 75x speedup, so it's totally
worth it. The downside is that queueing requires json,
but only on the server side.
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
| |
This provides the class-method behaviour that Ruby's JSON
support expects but that we don't provide.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Also making some log messages more informative.
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Up to now, when trying to match with tags:
File<<| tag == 'value' |>>
in fact we were querying parameters. Hopefully all the user tags
are stored in parameters so it was working.
But it wasn't possible to search on auto-tags (like class name).
This patch makes sure searching by tag is done on tags both on the
rails side and the resource side.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
|
|
| |
It appears that resources were not tagged with user tag on the
server, which prevents those tags to be persisted as tag in
the storeconfigs.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
I don't know why we imposed the restriction that we shouldn't match
with parameter containing arrays in exported mode.
That doesn't seem right, as the produced rails query works fine with
arrays.
Note: the user tags are not stored in the rails database except under
the special resource parameter tag. This also doesn't seem right.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Mongrel puppet code uses REMOTE_ADDR to set the ip address which will
be use to authenticate the client access.
Since mongrel is always used in a proxy mode with Puppet, REMOTE_ADDR
is always the address of the proxy (usually 127.0.0.1), which defeats
the purpose.
With this changeset, the mongrel code now uses the X-Forwarded-For
HTTP header value if it is passed over the REMOTE_ADDR.
Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
|
|
|
|
|
|
| |
- rename some tests to make their purpose clearer
- add a test for nested predicates
- remove trailing whitespace
|
| |
|
|
|
|
|
|
| |
Change the syntax for match in an onlyif back to using '==' for equality
comparison instead of 'eq' (this was the behavior before 9a2642) and make
operator for inequality '!='.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This includes reworking the get and match commands as well. This change
introduces a few small changes. These are:
1) There can be no ' or " characters around path elements.
2) The context is now only prepended to relative paths. So, if a
path is specified as /foo/bar the context will not be appended. It
will only be appended if it is foo/bar
3) The syntax for array matching in the onlyif is now eq or noteq.
It was == before.
4) The get and set commnands used to concatenate all items at the end
of a string so "set path to some value" would be interpreted as
"set" "path" "to some value". This is no longer supported. The caller
must put ' or " around the "to some value" for it to work"
Cherry-pick of 9a2642 from 0.24.x
Conflicts:
lib/puppet/provider/augeas/augeas.rb
|
|
|
|
|
|
| |
This commit should have no functional effect.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
|
|
| |
We often didn't set a value, unless it was true, which
meant that if it had previously been true but was now
false, we didn't fix it.
We also were not always saving modified resources, which
in some cases resulted in work not getting saved.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
| |
This somehow got lost in the conversion from
Parser resources to Puppet resources. We now
copy over the 'exported' value.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
| |
This allows a more fine-grained load-balancing
of the queue, which makes it easy to spin up
multiple puppetqd instances and process the
queues faster.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
requires certificate serial number to be strictly positive.
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
|
|
| |
This method name clashed with Puppet::Indirector::Envelope#expired?,
and its name wasn't actually very appropriate.
The new method name is 'dependent_data_expired?'.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
| |
This is just a simple call to Puppet.use in the puppetd
setup method.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
| |
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
| |
We weren't correctly propagating options through
to the Configurer instance.
Signed-off-by: Luke Kanies <luke@madstop.com>
|
|
|
|
|
|
|
| |
All of the settings were there, we just weren't doing anything
about them.
Signed-off-by: Luke Kanies <luke@madstop.com>
|