Jira (PUP-3696) Settings in puppet.conf not available in puppet itself
Title: Message Title Damon Atkins created an issue Puppet / PUP-3696 Settings in puppet.conf not available in puppet itself Issue Type: Bug Affects Versions: PUP 3.7.3 Assignee: Unassigned Components: Puppet Server Created: 2014/11/24 8:18 PM Environment: All Labels: puppetdb Priority: Normal Reporter: Damon Atkins Here is an example from puppetdb module. It hard codes the location of puppet configuration directory. ./params.pp: $puppet_confdir = '/etc/puppetlabs/puppet' ./params.pp: $puppet_confdir = '/etc/puppet' ./params.pp: $puppet_conf = "$ {puppet_confdir} /puppet.conf"
Jira (PUP-3695) yum provider handling for --disableexcludes in :install_options
Title: Message Title Joshua Hoblitt updated an issue Puppet / PUP-3695 yum provider handling for --disableexcludes in :install_options Change By: Joshua Hoblitt Summary: yum provider handling for -- disablexcludes disableexcludes in :install_options Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3695) yum provider handling for --disablexcludes in :install_options
Title: Message Title Joshua Hoblitt created an issue Puppet / PUP-3695 yum provider handling for --disablexcludes in :install_options Issue Type: New Feature Assignee: Kylo Ginsberg Components: Types and Providers Created: 2014/11/24 6:30 PM Priority: Normal Reporter: Joshua Hoblitt The yum provider has special handling for the --enablerepo and --disablrepo flags, if present in :install_options. The --disablexecludes option requires similar support. However, with the removal of yumhelper.py, it may be possible to pass :install_options verbatium to both `yum install` and `yum check-update` and remove the special handling of options completely. Add Comment
Jira (PUP-1793) Cleanup, reduce, and improve stdlib based on new Puppet 4 features
Title: Message Title Hunter Haugen commented on PUP-1793 Re: Cleanup, reduce, and improve stdlib based on new Puppet 4 features An update before heading off for Thanksgiving vacation: (caveat, I haven't read Henrick's 800-line write up yet, but it looks super useful and will go through it after vacation.) The high-level plan is to release a major version that drops puppet 2.7 support and adds puppet 4 support in a following minor release if not at the same time) so that we can have overlapping 3/4 support in a major revision of stdlib. Splitting 3/4 support between two modules would be a Bad Thing for our customers. This will cause some conflicting functions (ie, type()) and some gaps in functionality as mentioned, but nothing we can't cover with a few minor releases and some hacky puppet version detection (like having type.rb wrap the function declaration). The major revision after that will be a long time after and drop puppet 3 support. We would like if at all possible to avoid any other "churn" on major versions to keep user confidence high. I'll construct a more specific plan in the near future. Thanks for doing the footwork on this! Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-2563) Remove 1.8.7 support in the ruby code base
Title: Message Title Kylo Ginsberg assigned an issue to Kylo Ginsberg Puppet / PUP-2563 Remove 1.8.7 support in the ruby code base Change By: Kylo Ginsberg Assignee: Kylo Ginsberg Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3016) tests/modules/install/with_version.rb assumes `puppet` is on the path
Title: Message Title Kurt Wall commented on PUP-3016 Re: tests/modules/install/with_version.rb assumes `puppet` is on the path @rob Is there a way to restart the shell from a Beaker test, so we can do something like this: # test #1 do_stuff_that_doesn't_need_puppet restart_shell do_stuff_that_requires_puppet exit In Linux, we'd just source .some_file to refresh the env vars and/or say export YES=NO at the command line. In extremis, exec $SHELL to get a fresh shell without having to exit. Does Windows have something comparable? Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e)
Jira (PUP-2144) yumhelper.py uses whichever python shows up in the path first
Title: Message Title Joshua Hoblitt commented on PUP-2144 Re: yumhelper.py uses whichever python shows up in the path first bah - accidental status change. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-2144) yumhelper.py uses whichever python shows up in the path first
Title: Message Title Joshua Hoblitt commented on PUP-2144 Re: yumhelper.py uses whichever python shows up in the path first Since PUP-1362 has been merged and yumhelper.py no longer exists on master, can this issue be closed? Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3274) Remove old parser/evaluator
Title: Message Title Joshua Partlow commented on PUP-3274 Re: Remove old parser/evaluator merged to master in 5ab432b Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1072) support HTTP(S) URL as the file 'source'
Title: Message Title Felix Frank commented on PUP-1072 Re: support HTTP(S) URL as the file 'source' The sketch of Anthony Ryan is intriguing. I suggest moving it to a separate ticket with the purpose of future improvement. To move this forward, I shall try and look at a KISS way of implementing the support inside the current structure. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3341) Puppet apply breaks when an ENC returns an environment
Title: Message Title Felix Frank updated an issue Puppet / PUP-3341 Puppet apply breaks when an ENC returns an environment Change By: Felix Frank Story Points: 1 2 Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1515) Invalid parameter provider for custom types/providers
Title: Message Title Joshua Partlow commented on PUP-1515 Re: Invalid parameter provider for custom types/providers Commented on the PR; Charlie's patch looks like the correct behavior to me. A test run of acceptance on rhel7 passed as well. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (FACT-755) some xen hosts are not marked as xen
Title: Message Title Peter Huene updated an issue Facter / FACT-755 some xen hosts are not marked as xen Change By: Peter Huene Labels: community Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3694) puppet cert fingerprint invalidhost returns zero instead of non-zero
Title: Message Title Teran McKinney updated an issue Puppet / PUP-3694 puppet cert fingerprint invalidhost returns zero instead of non-zero Change By: Teran McKinney {{ # puppet cert fingerprint notfound; echo $?Error: Could not find certificate for notfound0 }} This should return 1 or another non-zero exist code since it does not exist. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3694) puppet cert fingerprint invalidhost returns zero instead of non-zero
Title: Message Title Teran McKinney created an issue Puppet / PUP-3694 puppet cert fingerprint invalidhost returns zero instead of non-zero Issue Type: Bug Assignee: Andy Parker Components: Server Created: 2014/11/24 3:32 PM Priority: Normal Reporter: Teran McKinney Original Estimate: 1 hour Remaining Estimate: 1 hour {{ puppet cert fingerprint notfound; echo $? Error: Could not find certificate for notfound 0 }} This should return 1 or another non-zero exist code since it does not exist.
Jira (PUP-3140) Illegal number error should include the faulty number
Title: Message Title Henrik Lindberg commented on PUP-3140 Re: Illegal number error should include the faulty number merged to master at f71f2f5 Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-659) Qualified variable lookups are very slow under Puppet 2.7+ unless prefixed with ::
Title: Message Title Henrik Lindberg commented on PUP-659 Re: Qualified variable lookups are very slow under Puppet 2.7+ unless prefixed with :: Since we have changed dynamic lookups to lexical for all variables, the search space is a lot smaller in 4.0. We have not measured the difference, and we have dropped support for Ruby 1.8.7. We should do a new measure at least. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1073) Common package name in two different providers
Title: Message Title Michael Smith assigned an issue to Michael Smith Puppet / PUP-1073 Common package name in two different providers Change By: Michael Smith Assignee: Michael Smith Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-2105) Puppet config face and the --configprint option operate directly on environment.conf.
Title: Message Title Nick Lewis commented on PUP-2105 Re: Puppet config face and the --configprint option operate directly on environment.conf. The use case in PE is for the agent installer to be able to set puppet.conf to point at the correct server before puppet runs. In that case, there's no module available that can do it. I don't have much opinion beyond wanting that specific use case to continue working. We don't have a specific need to modify environment.conf. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-2105) Puppet config face and the --configprint option operate directly on environment.conf.
Title: Message Title Eric Sorenson commented on PUP-2105 Re: Puppet config face and the --configprint option operate directly on environment.conf. Thinking on this a little more deeply, I'm not sure we need puppet to programmatically manipulate these files with the subcommand. There are puppet modules to manipulate line-oriented config files like this, and indeed afaict this is what the PE installer uses (although this feature was originally added for the installer!). [~nlew] J.D. Welch any thoughts? Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3389) Significant delay in puppet runs with growing numbers of directory environments
Title: Message Title Joshua Partlow commented on PUP-3389 Re: Significant delay in puppet runs with growing numbers of directory environments This is a problem in the environment loaders, in that the Directories loader when responding to a get request naively obtains a list of valid_directories() transformed into Node::Environment instances, and then returns the matching one. Each Node::Environment instance it creates, though, requires a Setting lookup for modulepath/manifest/config_version, and that call requires that settings find the environment.conf, which comes back around to the Directories loader. It's get_conf() method is also naively relying on valid_directories() listing all the directories, and then loading the environment.conf from the one that matches. So you get num_env^2 stats (* two because this whole process happens once for the initial autoload search for available_application_names, and then again (on the agent) when using the settings catalog) (+ some change for stating the base environmentpath dir each time). The master case is similar, though there it's push_application_context. After we've gotten past all this horribleness, there's a cached environment and we can proceed. On the master, a request in an environment that is not yet cached will incur num_env stats *2 (+ change); once for the initial lookup of valid_directories for the env, and then once again because the cache loader checks environment.conf to get ttl. We don't end up with the square here because settings are cached per environment. Anyway, the principle thing here is that the Directories loader be made a little smarter, so that get and get_conf do not rely directly on list and valid_directories. I think I can take care of this without too much trouble. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this gro
Jira (PUP-3389) Significant delay in puppet runs with growing numbers of directory environments
Title: Message Title Joshua Partlow updated an issue Puppet / PUP-3389 Significant delay in puppet runs with growing numbers of directory environments Change By: Joshua Partlow Story Points: 3 2 Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1802) Puppet should execute ruby.exe not cmd.exe when running as a windows service
Title: Message Title Michael Smith commented on PUP-1802 Re: Puppet should execute ruby.exe not cmd.exe when running as a windows service Merged https://github.com/puppetlabs/puppet_for_the_win/commit/2476cd007b1b52fb1b9f22202b56218629626b5c Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3389) Significant delay in puppet runs with growing numbers of directory environments
Title: Message Title Joshua Partlow updated an issue Puppet / PUP-3389 Significant delay in puppet runs with growing numbers of directory environments Change By: Joshua Partlow Story Points: 3 Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1793) Cleanup, reduce, and improve stdlib based on new Puppet 4 features
Title: Message Title Alexander Johnson commented on PUP-1793 Re: Cleanup, reduce, and improve stdlib based on new Puppet 4 features Hunter Haugen Can you please take a look at this and provide your opinion as the best way to move forward. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1793) Cleanup, reduce, and improve stdlib based on new Puppet 4 features
Title: Message Title Henrik Lindberg commented on PUP-1793 Re: Cleanup, reduce, and improve stdlib based on new Puppet 4 features At minimum the functions that use keywords (i.e. "type", and "private" must be changed to get a standardlib where thos (renamed) functions can be used. The approaches suggested above does not work for them. If we just ignore the problem, people cannot call those functions in 4.0 as they will be getting syntax errors. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-1793) Cleanup, reduce, and improve stdlib based on new Puppet 4 features
Title: Message Title Henrik Lindberg commented on PUP-1793 Re: Cleanup, reduce, and improve stdlib based on new Puppet 4 features Sine the review was made, there is now also a function called "private" that clashes with 4.0 keywords. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3656) Prefetch eats all exceptions, there is no way to fail out.
Title: Message Title Matt Carroll commented on PUP-3656 Re: Prefetch eats all exceptions, there is no way to fail out. Just a suggestion here, I think that to create an exception type which will be raised instead of logged on this level if it's rescued would solve this problem - that way provider authors can raise this exception type in the specific case that a failure means that a run should stop. This approach has the benefit that it introduces the required functionality without breaking any old behaviour, and puts the logic for what exceptions should be re-raised as fatal on the provider level. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3693) Adding star in value field of classifier causes html error
Title: Message Title Gerrard Geldenhuis created an issue Puppet / PUP-3693 Adding star in value field of classifier causes html error Issue Type: Bug Assignee: Unassigned Attachments: Screen Shot 2014-11-24 at 15.33.30.png Created: 2014/11/24 7:42 AM Priority: Normal Reporter: Gerrard Geldenhuis Screenshot should explain it all. Latest PE 3.7.0 on CentOS 6.5 using latest Google Chrome browser Add Comment
Jira (PUP-3692) Resource defaults leak between classes at the same level of scope hierarchy
Title: Message Title Henrik Lindberg commented on PUP-3692 Re: Resource defaults leak between classes at the same level of scope hierarchy The Resource Defaults are indeed messy, but they work according to how they were once designed to work (so not a bug). For the 4.0.0 release we tried to do something about them ( PUP-867 ) by making them apply to lexical scope instead of "dynamic scope". When we however changed this, we basically ended up with a useless defaults _expression_ and we broke every design that relied on resource defaults beyond trivial cases of not having to repeat resource attributes in a manifest. That change was therefore reverted (sad panda). After the 4.0.0 release we are going to address catalog building ("compilation" - which the application of default attributes are part of). We have several requirements to change the semantics of resource definitions, users want to be able to state the same resource multiple times and merge attributes if not conflicting and to be given a mechanism to override such conflicts if they arise. In order for this to be of value in practice, the scoping rules must involve modules and their dependencies - if not, all defaults and configurations bubble up to the top "node" level which is bad when trying to define a modular system. There must also be a mechanism to override defaults in a predictable way; now, you either have to carefully design the resource defaults expressions (to make them apply in an order that gives the wanted result), or perform collect/overrides to "patch" those resources that have undesirable values set. (These work arounds are certainly not appealing as they just hide the spaghetti with a layer of ravioli...) We have not yet started on the implementation of the new catalog builder, but we have ideas about how it should work that we plan to write up, present and discuss. Meanwhile, in 3.7 future parser and in 4.0.0, there are new features in the resource instantiation _expression_ that allow default values to be specified. These are of value to avoid "mysterious attribute values set by default" as they have higher precedence than the Resource Default expressions set per resource type. Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e)
Jira (PUP-3692) Resource defaults leak between classes at the same level of scope hierarchy
Title: Message Title Henrik Lindberg updated an issue Puppet / PUP-3692 Resource defaults leak between classes at the same level of scope hierarchy Change By: Henrik Lindberg Issue Type: Bug Improvement Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3692) Resource defaults leak between classes at the same level of scope hierarchy
Title: Message Title Geoff Williams commented on PUP-3692 Re: Resource defaults leak between classes at the same level of scope hierarchy moreinfo supplied by Gary Larizza Add Comment This message was sent by Atlassian JIRA (v6.3.7#6337-sha1:2ed701e) -- You received this message because you are subscribed to the Google Groups "Puppet Bugs" group. To unsubscribe from this group and stop receiving emails from it, send an email to puppet-bugs+unsubscr...@googlegroups.com. To post to this group, send email to puppet-bugs@googlegroups.com. Visit this group at http://groups.google.com/group/puppet-bugs. For more options, visit https://groups.google.com/d/optout.
Jira (PUP-3692) Resource defaults leak between classes at the same level of scope hierarchy
Title: Message Title Geoff Williams updated an issue Puppet / PUP-3692 Resource defaults leak between classes at the same level of scope hierarchy Change By: Geoff Williams Resource defaults set in 'class' scope currently leak into adjacent classes taking effect in the order they are parsed. I believe this is known internally as "dynamic scoping".As the use of puppet forge and modules in general has started to increase, this is starting to cause problems that are hard to diagnose and fix.This problem typically manifests itself as incorrect file ownership due to setting resource defaults for the File resource, but of-course could affect any resource type in puppet.h3. Problems this causes*Breaks idempotency of include function*The parse order of include statements changes the effect of the final catalogue - see testcase*Resource defaults leak into unrelated classes*Depending on parse order, its possible for resource defaults to leak into unrelated classes and alter the effect of them in subtle ways - in the past this has broken the concat module to the extent that puppet fails to run (see linked issues). I'm sure there are other examples of this too.*Hard to fix modules elegantly*Unwanted resource defaults are not easy to fix elegantly as they can require invasive code changes to protect against external influences. I have personal experience of this with the linked bug report to the concat module. In this case, you could argue the 'module above/including us' is the one with the problem since concat normally works fine. It becomes hard to see elegant ways to fix problems or indeed which module has a problem in the first place - the including one? the included one? both? puppet itself? See link to discussion on hunner-wordpress module's github issues page.*Against the puppet ideology*IMO puppet is supposed to be idempotent, with transparent and intuative results. The "dynamic scoping" of resource defaults goes against these principles.h3. TestcaseThe attached testcase has two test files that should be funtionally identical but are not, because they include classes in different orders. This results in different file ownership for {{/tmp/defaults_testcase}} file, and worse, the {{/tmp/unrelated}} file which picks up the unwanted resource defaults even thought its in another module:{noformat}[root@master /etc/puppetlabs/puppet]# puppet apply modules/defaults_testcase/tests/order1.ppNotice: Compiled catalog for master.puppetlabs.vm in environment production in 0.11 secondsNotice: /Stage[main]/Defaults_testcase::Class_c/File[/tmp/defaults_testcase]/ensure: createdNotice: /Stage[main]/Unrelated/File[/tmp/unrelated]/ensure: createdNotice: Finished catalog run in 0.15 seconds[root@master /etc/puppetlabs/puppet]# ls -l /tmp/defaults_testcase /tmp/unrelated -rw-r--r-- 1 wordpress wordpress 0 Nov 24 07:28 /tmp/defaults_testcase-rw-r--r-- 1 wordpress wordpress 0 Nov 24 07:28 /tmp/unrelated[root@master /etc/puppetlabs/puppet]# puppet apply modules/defaults_testcase/tests/order2.ppNotice: Compiled catalog for master.puppetlabs.vm in environment production in 0.12 secondsNotice: /Stage[main]/Defaults_testcase::Class_c/File[/tmp/defaults_testcase]/owner: owner changed 'wordpress' to 'drupal'Notice: /Stage[main]/Defaults_testcase::Class_c/File[/tmp/defaults