summaryrefslogtreecommitdiffstats
path: root/lib/puppet/indirector
Commit message (Collapse)AuthorAgeFilesLines
* Merge branch '0.25.x'Markus Roberts2010-02-092-0/+4
|\ | | | | | | | | | | | | | | | | | | Conflicts: lib/puppet/agent.rb lib/puppet/application/puppet.rb lib/puppet/configurer.rb man/man5/puppet.conf.5 spec/integration/defaults.rb spec/unit/configurer.rb
| * Partial reversion of patch for #3088 to fix #3104 (Exception misreported)Markus Roberts2010-01-242-2/+2
| | | | | | | | | | | | | | In my patch for #3088 I made a erroneous assumption about the ruby exception hierarchy and thus missed the fact that Timeout::error descends from both SignalError and Interrupt. This is a partial reversion of the patch for #3088 to let these through so that more useful error messages can be produced.
| * Fix for #3088 (catching Exception also traps SystemExit)Markus Roberts2010-01-242-0/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | Changing rescues from the default to Exception (to catch errors that don't descend from StandardError) had the unintended consequence of catching (and suppressing) SystemExit. This patch restores the behavior of by reraising the exception. Of the other exceptions that fall through the same crack (NoMemoryError, SignalException, LoadError, Interrupt, NotImplementedError, and ScriptError) this patch also reraises NoMemoryError, SignalException, and Interrupt in the same way and leaves the rest captured.
* | Merge branch '0.25.x'James Turnbull2010-01-134-23/+21
|\| | | | | | | | | | | Conflicts: lib/puppet/ssl/host.rb spec/spec_helper.rb
| * Fix for #2765 (--no-fqdn regression in puppetrun)Markus Roberts2009-12-201-3/+6
| | | | | | | | | | | | | | | | | | This is basically the fix suggested on the ticket, cleaned up and ruby-ized, with tests. The only functional modification is leaving the default on entry2hash as --no-fqdn to preserve 0.25.1 behaviour as the default. Signed- ff-by: Markus Roberts <Markus@reality.com>
| * Fix for #2890 (the cached certificates that would not die)Markus Roberts2009-12-193-20/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch implements the two-part suggestion from the ticket; 1) a client that receives a certificate that doesn't match its current private key does not accept, store or use the certificate--instead it removes any locally cached copies and acts as if the certificate had never been found. 2) a puppetmaster that receives a csr from a client for whom it already has a signed certificate now honors the request and considers it to supercede any previously signed certificates. In order to make the cache expiration work as expected, I changed a few assumptions in the caching system: * The expiration of a cached certificate is the earlier of the envelope expiration and the certificate's expiration, as opposed to just overriding the cache value * Telling the cache to expire an item now removes it from the cache if possible, rather than just setting an expiration date in the past and hoping that somebody notices. Signed-off-by: Markus Roberts <Markus@reality.com>
* | Merge branch '0.25.x'Luke Kanies2009-12-214-3/+8
|\| | | | | | | | | | | | | | | Conflicts: lib/puppet/agent.rb lib/puppet/application/puppetd.rb lib/puppet/parser/ast/leaf.rb lib/puppet/util/rdoc/parser.rb
| * Possible workaround for #2824 (MRI GC bug)Markus Roberts2009-11-191-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a moderately ugly workaround for the MRI garbage collection bug (see the ticket for details). I explored several other potential solutions (notably, monkey patching the routines that trigger the bug) but none of them were satisfactory. Monkey patching sub, gsub, sub!, gsub!, etc., for example, either changes the scoping of $~, $1, etc. in a way that could potentially subtly change the meaning of programs or (if you are clever) faithfully reproduces the behaviour of MRI--including the memory leak. I decided to go with the standardized and somewhat obnoxious never- used optional argument as it was easy to automatically insert and should be even easier to automatically find and remove if a better fix is developed. It also should be obtrusive enough to escape accidental removal in refactoring.
| * Fixed ldap typo againJames Turnbull2009-11-151-1/+1
| |
| * Fix for #2661 and related issuesMarkus Roberts2009-11-131-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If setup code for a process depends on network connectivity it needs to be protected with a rescue clause as much as the main body of the process. Further, Timeout exceptions aren't under StandardError and thus aren't caught by an un-typed rescue clause. This doesn't matter if we've morphed the exception, but will cause the program to fail if we haven't. There are many places where these concerns _might_ cause a problem but in most cases they never will in practice; this patch addresses the five cases where I have been able to confirm that it actually can cause the client daemon to exit and two more where I suspect (but can not prove) that it could. This is an extension of the prior patch to cover additional cases found by automated testing (repeated catalog runs with a 1% chance of timeout forced on all timeout-bound operations, ~5000 runs). The new cases recurred multiple times (>100 each) and in a final pass with these corrected (~2500 runs) no additional cases were found.
| * Fix #2787 - Storeconfig doesn't store/update node ip and environment anymoreBrice Figureau2009-11-071-0/+5
| | | | | | | | | | | | | | | | | | Since the storeconfig refactoring (ie moving the catalog storeconfig system under the indirector) in 0.25 we lost the capability to store the node ip and node environment name. This patch restores this feature. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
| * Fixed typo in lib/puppet/indirector/ldap.rbJames Turnbull2009-10-271-1/+1
| |
| * Revised partial fix for #2661 and related issuesMarkus Roberts2009-10-272-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If setup code for a process depends on network connectivity it needs to be protected with a rescue clause as much as the main body of the process. Further, Timeout exceptions aren't under StandardError and thus aren't caught by an un-typed rescue clause. This doesn't matter if we've morphed the exception, but will cause the program to fail if we haven't. There are many places where these concerns _might_ cause a problem but in most cases they never will in practice; this patch addesses the two cases where I have been able to confirm that it actually can cause the client daemon to exit and two more where I suspect (but can not prove) that it could. I'd be willing to push this patch as it stands, as it at least fixes demonstrable problems. A more general solution would be nice.
* | Fixing #2617 - using the searched-for REST nameLuke Kanies2009-11-201-1/+3
|/ | | | | | | | | This allows a separation between the wrapper class and its internals, which is (at least) necessary for the CA cert, which might not be found using the internal name. Signed-off-by: Luke Kanies <luke@madstop.com>
* Bundling of pure ruby json lib as "pson"Markus Roberts2009-10-171-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Bundeling and renaming the pure ruby json library to addresses a number of cross version serliaization bugs (#2615, et al). This patch adds a subset of the files from the json_pure gem to lib/puppet/external/pson (renamed to avoid conflicts with rails) so that we will always have a known-good erialization format available. The pure ruby json gem as distibuted defers to the compiled version if it is installed. This is problematic in some circumstances so the files that have been brought over have been modified to always and only use the bundled version. It's a large patch, so here's a breakdown of the change categories: The majority of the lines are only marginally interesting: * The json lib itself (in lib/puppet/external/pson) make up the bulk of the lines. * Renaming of json to pson make up the second largest group. Somewhat more interesting are the following, which can be located by searching the diffs for the indicated strings: * Adjusting tests to reflect the changes * Changing the encoding/decoding behavior so that nested structures (e.g. resources) don't serialize as escaped strings. This should make it much easier to process the results with external tools, if needed. Search for "to_pson" and "to_pson_data_hash" * Cleaning up the envelope/metadata * Now provides a document_type (as opposed to a ruby class name) by using a symple registration scheme instead of constant lookup (search for "document_type") * Added an api_version (search for "api_version") * Added a hash for document metadata (search for "metadata") * Removing the yaml monkeypatch and instead disabling yaml serialization on ruby 1.8.1 in favor of pson (search for "yaml") * Cleaning up the json/rails feature interaction (they're now totally independent) (search for "feature")
* Fixing #2617 - use the cert name as specifiedLuke Kanies2009-09-221-0/+6
| | | | | | | | | | This allows us to search for a cert, and we use the searched-for term as the cert name (for the wrapper, not the actual cert object), rather than the real cert name. This allows us to use symbolic names like 'ca', as we're currently doing. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fix for #2600 (wrong number of arguments under older mongrel)Markus Roberts2009-09-151-1/+1
| | | | | | | | | | | | | | | | | | This was an API compatibility problem with mongrel's HTTPResponse.start() method between Mongrel 1.0.x and 1.1.x (the number of parameters changed). The older version does not provide the option to set the response header message which was used (redundantly with the response body) to return the error message when the HTTP response was signaling an error. In order to suport the older version the call was wrapped with a fallback and the coresponding code in the other rest implementations was adjusted to always send the error message in the response body. Then the rest terminus was adjusted to pull the message from the response body (if it is present) rather than from the header (which is only used as a fallback for dealing with older puppetmasters), and the tests were augmeted to verify this behaviour. Signed-off-by: Markus Roberts <Markus@reality.com>
* Update documentation string to reflect actual intent of ↵Steven Jenkins2009-09-041-1/+1
| | | | Puppet::Node::Facts::Rest
* Fixing #2558 - propagating recent fileserving changesLuke Kanies2009-08-241-2/+2
| | | | | | | | | | | | | | | | | | I'd made changes to the internals of the fileserving system to fix #2544 (mostly switched from passing the node around and then calculating the environment to just passing the environment around), but those changes weren't consistent throughout the fileserving code. In the process of making them consistent, I realized that the plain file server actually needs the node name rather than the environment, so I switched to passing the request around, because it has both pieces of information. Also added further integration tests which will hopefully keep this from cropping up again. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing #1544 - plugins in modules now works againLuke Kanies2009-08-181-4/+5
| | | | | | | | | | | | We had to fix the fileserving plumbing to use the request environment instead of trying to use the node environment. This was apparently never fixed after we added the environment to the URI in REST calls. There's still a bit of refactoring left to clean up the APIs used in some of this code. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing fact-missing problem when puppet.conf is reparsedLuke Kanies2009-08-181-0/+6
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Fix #2516 - Fix format detection when content-type contains charsetBrice Figureau2009-08-131-2/+4
| | | | | | | | | | | | | | | | | | Even though Puppet never transmist a charset information in its response/request content-type, some proxy (especially Apache with the infamous AddDefaultCharset configuration) may add this "incorrect" information. This patch makes sure that only the mime-type is used when looking for the format associated with a response or a request. The patch also provides a better error message when the client or server code is fed with a request whose mime-type can not be mapped to a known format. It also fixes a typo noticed by the original reporter. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2507 - Exported resources were not correctly collected.Brice Figureau2009-08-131-1/+1
| | | | | | | | | | | | | | | | | | | | | | #2507 contains two issues: * a crash when we filters-out an unwanted resource which had edges pointing to it. * resources are losing their virtuality when they are transformed from Puppet::Parser::Resource to Puppet::Resource. This means we weren't able to distinguish anymore between an exported resource collected in the same node as it was exported and an exported resource collected in another node. The net result is that we can't apply exported resources that are collected in the same node because they are filtered out by the catalog filter (see the commits for #2391 for more information). The fix is to keep the virtuality of the resources so that we can differentiate those two types of exported resources. We keep this until the catalog is ready to be sent, where we filter out the virtual resouces only, the other still exported ones needs to be sent to the client. To be real sure, the transaction also skips virtual resources. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fixing #2489 - queue integration tests are skipped w/out jsonLuke Kanies2009-08-101-1/+0
| | | | | | | | | All of the tests were failing because we had a call outside of any of the tests, just to autoload the constant. Removed that call and stubbed things so the tests don't run without json. Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixes #2464, #2457. Deprecate reportserver for report_server. Add ↵Nigel Kersten2009-08-041-0/+2
| | | | report_port setting. Add tests.
* Modifying the REST client error to make server errors more clearLuke Kanies2009-08-011-1/+1
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Fix #2425 - make sure client can contact CA server with RESTBrice Figureau2009-07-213-0/+9
| | | | | | | The various REST SSL terminii were never setup to use the ca_server/ca_port if one is setup. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2424 - File server can't find module in environmentBrice Figureau2009-07-181-2/+2
| | | | | | | | | | | | | | | | | | | | | | | Actually, the issue is: * when the web server gets the request, it creates an indirection request, filling attributes like ip or node from the HTTP request. To do this, all the interesting attributes are given in a hash (called options, see P::I::Request#new). Once the request is properly initialized the options hash doesn't contain the ip or node information (see set_attributes) * the request is then transmitted to the file_serving layer, which happily wants to use the node attribute to find environments or perform authorization. Unfortunately it fetches the node value from the request options hash, not the request itself. Since this node information is empty, puppet fails to find the proper mount point, and fails the download. This change makes sure we pass all the way down the node and fix the authorization check. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2391 - Exported resources never make to the storeconfigs dbBrice Figureau2009-07-182-1/+7
| | | | | | | | | | | | | | | | | | The issue is that when we convert Puppet::Parser::Resource catalog to a Puppet::Resource catalog before storing it to the database, we don't allow virtual resource to be converted. Unfortunately exported resources are virtual by design, and as such aren't converted, and we lose them, so it isn't possible to store them in the database. Unfortunately, the client will get the exported resources too. The fix is dual-fold: * we make sure exported resource are skipped when the transaction is applied as a last safeguard * we filter-out the catalog through the catalog compiler terminus before the catalog is returned to the client Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2261 - Make sure query string parameters are properly escapedBrice Figureau2009-07-171-4/+5
| | | | | | | | | | | | | | | | The problem is that URI.escape by default doesn't escape '+' (and some other characters). But some web framework (at least webrick) unescape the query string behind Puppet's back changing all '+' to spaces corrupting facts containing '+' characters (like base64 encoded values). The current fix makes sure we use CGI.escape for all query string parameters. Indirection keys/path are still using URI escaping because this part of the URI format shouldn't be handled like query string parameters (otherwise '/' url separators are encoded which changes the uri path). Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Fix #2392 - use Content-Type for REST communicationBrice Figureau2009-07-051-1/+1
| | | | | | | | | | | | | | | | | | | There were two problems: * server->client communications is using Content-Type with the direct format name instead of the format mime-type. * client->server communications is not using Content-Type to send the format of the serialized object. Instead it is using the first member of the Accept header. The Accept header is usually reserved for the other side, ie what the client will accept when the server will respond. This patch makes sure s->c communication contains correct Content-Type headers. This patch also adds a Content-Type header containing the mime-type of the object sent by the client when saving. Signed-off-by: Brice Figureau <brice-puppet@daysofwonder.com>
* Switching Queueing to using JSON instead of YAMLLuke Kanies2009-06-061-13/+20
| | | | | | 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.
* Removed extra whitespace from end of linesIan Taylor2009-06-0611-20/+20
|
* Fixing #2315 - ca --generate works againLuke Kanies2009-06-031-9/+1
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Changing Puppet::Cacher::Expirer#expired? method nameLuke Kanies2009-05-281-0/+1
| | | | | | | | | 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>
* Fixing #2240 - external node failures now log outputLuke Kanies2009-05-281-5/+1
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixes #2209 - Spec is failing due to a missing requireStéphan Gorget2009-05-021-0/+1
|
* Fixing #2149 - Facts are passed as part of the catalog requestLuke Kanies2009-04-221-9/+35
| | | | | | | | | | | | | | | This removes the requirement of shared fact caching on the servers, since the server responding to the catalog request will receive the facts as part of the request. The facts are serialized as a parameter to the request, rather than each being set as a separate request parameter. This hard-codes yaml as the serialization format for the facts, because I couldn't get marshal to work and it's just not as big a deal for such a small amount of data. Signed-off-by: Luke Kanies <luke@madstop.com>
* Adding time debugging for catalog storage to active_recordLuke Kanies2009-04-221-3/+5
| | | | Signed-off-by: Luke Kanies <luke@madstop.com>
* Using a setting for configuring queueingLuke Kanies2009-04-221-1/+0
| | | | | | | Removing class methods and such, and switching to relying solely on a setting for the queue configuration. Signed-off-by: Luke Kanies <luke@madstop.com>
* Removing unnecessary parser variables when yaml-dumpingLuke Kanies2009-04-222-5/+5
| | | | | | | | | Also working around a YAML bug in Ruby. And fixing tests that were broken in a previous commit in this rebase but not caught earlier. Signed-off-by: Luke Kanies <luke@madstop.com>
* Add queue indirection as an option for catalog storage.Ethan Rowe2009-04-221-0/+5
| | | | | | This eventually will allow catalog storage ("storeconfigs") to be taken out of the critical request-handling path of puppetmasterd, such that: * Puppet::Node::Catalog can be serialized to a message queue via the indirector's "save" method * a separate process can use Puppet::Node::Catalog::Queue.subscribe to pick up these catalog objects as they come in and can save them to the database through the :active_record terminus
* Introduce abstract queue terminus within the indirection system.Ethan Rowe2009-04-221-0/+78
| | | | The queue abstract terminus allows the standard indirector behaviors to interact with a message queue broker, such that the indirector's "save" method writes the relevant model object out to a queue on the message broker. While the indirector's "find" method does not map to a message queue, the queue terminus class offers a "subscribe" method that allows for easy implementation of an event loop, receiving indirected objects saved to a queue as they come in.
* Refactoring the Rails integrationLuke Kanies2009-04-222-2/+2
| | | | | | | | This moves all code from the Parser class into the ActiveRecord classes, and gets rid of 'ar_hash_merge'. Signed-off-by: Luke Kanies <luke@madstop.com>
* Adding ActiveRecord terminus classes for CatalogLuke Kanies2009-04-221-0/+34
| | | | | | | | This provides the last piece of ActiveRecord integration. It's pretty much just pass-through and *only* works if you're storing Parser resources to the db. Signed-off-by: Luke Kanies <luke@madstop.com>
* Adding ActiveRecord terminus classes for Node and Facts.Luke Kanies2009-04-223-0/+70
| | | | | | | | | This is most of the way to replacing standard StoreConfigs integration with the Indirector. We still need to convert the Catalog and then change all of the integraiton points (which is mostly the 'store' call in the Compiler). Signed-off-by: Luke Kanies <luke@madstop.com>
* Fixing #2171 - All certificate files are written with default permsLuke Kanies2009-04-222-5/+7
| | | | | | | | | | We were already writing all specific files with appropriate permissions; this change makes all of the files that are part of a group (which largely means files saved by puppetmasterd and puppetca) are also written using the correct permissions. Signed-off-by: Luke Kanies <luke@madstop.com>
* Partially fixing #2029 - failed caches doesn't throw an exceptionLuke Kanies2009-04-171-7/+17
| | | | | | | If the main terminus fails you get an exception, but not if a cache terminus fails. Signed-off-by: Luke Kanies <luke@madstop.com>
* Merge branch '0.24.x'Luke Kanies2009-04-021-39/+75
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Conflicts: bin/ralsh lib/puppet/executables/client/certhandler.rb lib/puppet/parser/functions/versioncmp.rb lib/puppet/parser/resource/reference.rb lib/puppet/provider/augeas/augeas.rb lib/puppet/provider/nameservice/directoryservice.rb lib/puppet/provider/ssh_authorized_key/parsed.rb lib/puppet/type.rb lib/puppet/type/file/checksum.rb spec/integration/defaults.rb spec/integration/transaction/report.rb spec/unit/executables/client/certhandler.rb spec/unit/indirector/ssl_rsa/file.rb spec/unit/node/catalog.rb spec/unit/provider/augeas/augeas.rb spec/unit/rails.rb spec/unit/type/ssh_authorized_key.rb spec/unit/type/tidy.rb test/executables/filebucket.rb test/executables/puppetbin.rb
| * Fixing #1991 - ldap booleans get converted to booleansLuke Kanies2009-03-051-0/+25
| | | | | | | | Signed-off-by: Luke Kanies <luke@madstop.com>
generated by cgit (git 2.34.1) at 2025-09-30 00:19:40 +0000