Activity
From 03/16/2011 to 03/22/2011
03/23/2011
-
10:47 pm Feature #6527 (Merged - Pending Release): Pip package provider
- Paired with Richard to cleanup a few things and merged into next commit:8542160479538fe6575475c2dccf5d4696fa3e5d
-
10:28 pm Facter Bug #5135 (In Topic Branch Pending Review): Facter returns physicalprocessorcount => 0 on a physi...
-
- Added timeout for prtdiag, corrected tests.
-
08:35 pm Bug #4487: Environment column in hosts table updating incorrectly
- I can reproduce at-will on 2.6.x and 2.6.next...
-
08:24 pm Bug #4487 (In Topic Branch Pending Review): Environment column in hosts table updating incorrectly
- I can reproduce this at-will on 2.6.x but not on 2.6.next.
<https://github.com/puppetlabs/puppet/compare/2.6.x...2... -
- Just poked at this for a while. I thought it might be due to us specially handling environment when parsing the confi...
-
07:02 pm Bug #6810: File 'fail' became a lot more fatal in 2.6.0
- Sorry, this goes back to 2.6.0. So the title is inaccurate. Is this an intentional change, then?
-
- OK, actually 2.6.4 behaves the same.
-
06:28 pm Bug #6814: Interface actions should be represented by an object
- When we implement this we should have two objects involved in the process:
class Action
attr_accessor... -
05:59 pm Bug #6814 (Accepted): Interface actions should be represented by an object
- This will allow us to define meta-data about an action, such as description and its options.
-
05:33 pm Bug #6786 (Accepted): Redundant mechanisms to access interface instances
- This seems to have not only broken some code (fixable), but to be a poor choice for the long term. I will be revisit...
-
05:32 pm Feature #5692: Internally-generated state files should be written in an atomic way
- Other ones I can think of:
* the new --summary output thats saved to disk
* the new feature to save a report to... -
- Vichharaks, can you confirm that the description of the issue above is actually the one you're hitting?
03/22/2011
-
04:58 pm Bug #3238: puppetmaster uses to many mysql connections
- I've just been bitten by the same problem. I've set wait_timeout=60 as a workaround but have not confirmed that fixes...
-
- Given:
a) Issues with Ruby's namespace lookup mechanism
b) The mental / namespace / other overhead of maintai... -
- Some OSes may not have a reasonable concept of a *default* admin user and admin group, like Windows.
You can't rel... -
01:49 pm Facter Feature #6801: Provide adminuser and admingroup facts for OSes where it makes sense.
- As I was reading the title of the ticket, I wondered about the "for OSes where it makes sense." Where doesn't it mak...
-
11:24 am Facter Feature #6682: Solaris LDOM facts
- My oracle support contact agree the best way to identify guest ldom domains is in this URL:
[http://opensolaris.org... -
08:24 am Bug #6562 (Merged - Pending Release): Minor fix in puppet kick help
- commit:a29c7fd408b8b1e00410f5bbd82ff5e626381667
-
- Looks like all you're doing is removing a blank comment line. Feel free to merge doc changes that trivial without re...
-
- I didn't notice this ticket until after I pushed the same fix as part of #6566. commit:e80063468e404a6827ea6ec458ae3...
-
- Committed to next along with a fix to a case statement that showed up when running `rake spec` after the futils thing...
-
- The puppet master still fails to start with these patches, it just gets a little farther. You get a slew of other er...
-
- You can get around this temporarily by just changing to the test dir, and then running rake. A longer term fix will ...
-
- Instead of the rake task you can run `rspec spec`, which will run the whole spec dir of tests and get around windows ...
-
04:19 am Feature #5528: API call to sign, clean or generate certificate request
- I'll do what I can to make sure the work we're doing fits this proposed API.
-
-
- In testing 2.6.6, we found that missing source directories in a recurse=>remote file copy cause puppetd to exit immed...
-
- A preliminary version of this is available in master as of commit:0f4a4300f18da04fc672475645c0d2ba9721d1d5.
-
- I updated the formatting to add pre tags.
-
- Here's an attempt at using disable only when sysv-rc package version is greater than or equal to 2.88, which seems to...
-
- Actually, does just adding:
links => follow
do the right thing for you? -
- James, not sure what code you updated.
Micah, formatting fixed for my patch with another patch coming that uses dpkg... -
-
- Code updated.
-
11:52 pm Bug #5908: Service Provider on Debian unable to disable services: now uses update-rc.d enable/dis...
- Speaking of bad formatting, this is the correct link to the [debian-devel thread](http://lists.debian.org/debian-deve...
-
- Matt Robinson wrote:
> If we depend on operating system version it could be a problem if somebody backports sysv-rc ... -
- We have been discussing this in the [debian bug](http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=573551) to try and ...
-
- Daniel Pittman wrote:
> It would be very nice to support the convention that `-` is an alias for stdin when you prov... -
- <https://github.com/puppetlabs/puppet-interfaces/tree/ticket/master/6774> is ready to be merged.
-
- If we depend on operating system version it could be a problem if somebody backports sysv-rc that has this problem.
... -
- It doesn't make sense to set "links => follow" inside a symlink resource. We're not managing content, and you end up ...
-
- ...
-
- It looks like this may be a bug with debian squeeze more than puppet. For some reason using update-rc.d service stop...
-
10:45 pm Feature #5660 (Tests Insufficient): Puppet should handle port entries in /etc/services with a nat...
- As posted on the mailing list:
Stefan,
I started taking a look at the patch series, but it doesn't ... -
- Implemented in 23f4ff338949e085e42d778809a97f39be021da9 and pushed.
-
- ...
-
-
- That would meet all of my needs.
-
- We don't do much error checking in the name2const and unify_name methods; they also don't well represent what that co...
-
- Yep. I'm positive this didn't actually work :)
http://projects.puppetlabs.com/issues/2585
Feel free to follow t... -
- We will be making versioning mandatory, as part of unification of the interface access paths.
The mechanism will b... -
-
09:17 pm Puppet Dashboard Feature #4052: arrays should support pre-pend and append overrides.
- The use case was resolvers, that as the groups get more specific, they may want to add another resolver as either the...
-
09:14 pm Bug #6804: +> operator only works on ignore in defaults context
- I can't say for certain that it actually worked before (everything has been upgraded now, so I don't have the setup i...
-
- Dan Bode wrote:
> I don't know if I can supply any additional information here. Order sometimes matters for arrays. ... -
- Are you sure it was actually working before? I thought the plusignment operator only ever worked in the context of cl...
-
- We should move away from the unix-centric "root" name here.
How about paired facts of:
* adminuser
* admin... -
- In my global site.pp, I have this default value set:...
-
-
- While most operating systems use root:*root* for ownership of *system* files, AIX uses system as the group by default...
-
-
-
- (#6799) is open for acceptance testing this fix.
-
- #6798 fixes a problem that requires communicating with a master, which is awkward if not impossible from a unit test....
-
- Not actually a bug: apparently I didn't have that installed in the ruby library path. Ahem.
-
- Fixed by preloading certificates: <https://github.com/puppetlabs/puppet-interfaces/tree/ticket/master/6798>
-
- The following command will cause infinite recursion:
puppet certificate find --from=rest $(hostname --fqdn)
... -
- Failure/Error: Puppet::Interface::Indirector.terminus_classes(:catalog).should be_include("compiler")
Coul... -
- Thanks, Todd. This has been merged into next as of commit:f7a87db74e9933e9de20489e30d640115ef36964
-
06:17 pm Feature #1886: Add node cleanup capability
- > 1. Are we shipping "interfaces", so that it would make sense to have 'puppet node clean' as an action?
> 2. Do we ...
03/21/2011
-
03:22 pm Facter Bug #6795 (In Topic Branch Pending Review): xendomains: Ignore error output from xm list
-
- If xend is not running the xendomains fact produces a spurious error from xm list (Error: Unable to connect to xend: ...
-
02:31 pm Facter Bug #6794 (Closed): Wrong uptime info
- Solaris 10 SPARC
Ruby 1.8.7p330 built and tested from source
Facter 1.5.8 installed from official gem
...
... -
01:56 pm Bug #6793 (Merged - Pending Release): mount provider fails when paths have a trailing slash
- A resource like
"/mnt/foo" works fine, but "/mnt/foo/" only works on the first mount. On subsequent runs, puppet in... -
08:05 am Bug #6744 (Closed): Multidimensional Array passed to template will be flattened
- Ok, this is our fault then. We have cloned the git repository because of some modifactions we made(zypper and yum pro...
-
- It would be very nice to support the convention that `-` is an alias for stdin when you provide a filename, or list o...
-
- I would really, really like to stop supporting YAML as an output format as much as possible. It has sufficiently hor...
-
05:21 am Bug #6748: ndefined method `<<' for {}:Hash when combining resource chaining, metaparameters, and...
- Is it worth renaming the ticket, to make it a little clearer of the issue?
-
-
- Are you working on this and tracking it Jeff or just reporting it and making the test?
-
03:58 am Feature #4258: Add pkgutil provider
- I've tested your changes, Dominic.
There was an issue with the hashes created for "aliases". I've send through a sma... -
- I would strongly suggest we make the version information *mandatory*, not optional, at the "interface" layer. If nec...
-
- 4.1.2.2. Serial Number
The serial number MUST be a positive integer assigned by the CA to
... -
- Luke Kanies wrote:
> Daniel Pittman wrote:
> >
> > I care much less about that, and much more about the fact that... -
- Daniel Pittman wrote:
>
> I care much less about that, and much more about the fact that these highly position-sen... -
- It's worth pointing out that while there's some amount of extra work here, I've also had consistent user feedback tha...
-
- I agree with Daniel - I agree that the current system is non-optimal, but I think your recommendation will miss probl...
-
- I'm comfortable with this as long as 'they have to be random' changes to 'they cannot collide'. It's true we've chos...
-
- I understand James Turnbull to be working on this.
03/20/2011
-
11:05 am Facter Feature #6792 (Closed): A fact to identify OS family
- Following the discusssion in ticket #6679 it could be a good thing to have a new fact that identifies the operatings...
-
- Just a side note.
I made my test on a SL6 BETA version so my operatingsystem fact evaluating to RedHat might have be... -
- Daniel Capo Sobral wrote:
> Daniel Pittman wrote:
> >
> > This only really matters for commands run through the e... -
11:57 pm Bug #4695: Puppet should not reset locale
- Daniel Pittman wrote:
>
> This only really matters for commands run through the exec type, I would imagine, so my ... -
07:30 pm Facter Bug #6679: Facter 1.5.8 does not identify Scientific Linux
- I have no problem if the consensus is to leave the redhat designation.
However, it feels incorrect to me. It seems t... -
- True. We went through this pain before when we moved from identifying Ubuntu as "debian" -> "ubuntu".
So... Is it ... -
- Jay Flowers wrote:
> The registry can be useful but is by no mean where you are supposed to put configuration. The ... -
- This will also break the maillist and user types on SL as the providers check for all the redhat like OS.
-
- I'm starting to wonder if we need an OS family fact of some kind. Feel free to open a feature request :)
There are... -
- This was introduced in 79dcd33aebf8749719e9eff009b8bb3fdaf77751, specifically to ensure that external tools run by Pu...
-
- It looks like the reports:prune task is only deleting reports, not destroying them. This causes the associated resour...
-
- This happens when there is a new-ish version (I haven't nailed down the particular version) of the i18n gem installed.
-
- This is because the active_record terminus that storeconfigs sets as the cache for facts does not properly return the...
03/19/2011
-
03:46 pm Bug #4695: Puppet should not reset locale
- sorry, patch in file :)
-
- This patch for debian.
I not know ruby, and i know - this is bad solution, but in debian's distr this work.... -
11:57 am Bug #6773: Install Script Fails; needs good default configdir for MS Windows
- The registry can be useful but is by no mean where you are supposed to put configuration. The standard for a .Net ap...
-
- I just updated Facter from 1.5.8 to 1.5.9 on a Scientific Linux 6 test box and found that the operatingsystem fact ch...
-
- Markus Roberts wrote:
> The workaround for this would be to specify the directory on the command line; e.g. somethin... -
04:16 am Bug #6772 (Closed): err: Could not run Puppet configuration client: Parameter require failed: No ...
- This is not an issue, I'm just unable to read english these days
Had
require => installed
Should be
ensure =>... -
01:08 am Feature #4258: Add pkgutil provider
- Updated method to handle package aliases (i.e. "subversion" is "CSWsvn") by calling pkgutil -a to translate between t...
-
-
11:24 pm Facter Bug #6763 (Merged - Pending Release): arp uses incorrect path for RHEL/CentOS installations
- Todd's patch merged into next in commit:b82635b
-
- The master exec now raises when there's a problem with the ENC script instead of logging. The result is that the err...
-
- If you're trying to manually create a node that already exists using the web ui there error you get doesn't tell you ...
-
- Merged into 2.6.next in commit:fa2010b
-
- After talking with James, it looks like propagating the block below down to the agent's run log would be sufficient f...
-
10:02 pm Feature #5400: Augeas resources should print changes to the tree during a --noop -t run
- This seems to be very similar to feature request #2728.
I've create a patch for #2728 which may also address this ... -
- Once we've got the interfaces work complete and merged in, we should migrate this class to use them, rather than stan...
-
- This has been bugging me for a while, so I've created a patch so Puppet will show the changes that the Augeas provide...
-
- There's some functionality crossover here with the cloud plans, particularly with an authenticated service for cloud ...
-
- We should provide the interface demonstrated on the last line of the ticket's body, but providing the options hash sh...
-
09:39 pm Feature #6749: Actions need to support options
- A simple API to support this would be to allow a list of optparse strings to be supplied as an argument to the "actio...
-
- Actions should be able to specify the default values to use for unspecified options.
Additionally, since an interf... -
- Interface instances may be accessed in two ways:
Puppet::Interface::interface(:<interface_name>)
Or:
... -
- In interfaces that are based on indirections, the user may select the terminus using the option "--from". This is co...
-
- "I believe this ticket will be rendered unnecessary by ticket #6781. Some of Dan’s logic may need to be moved from th...
-
- Yes this was fixed in one of the Facter 1.5.x releases.
-
- The code currently implements Interface creation as:
Puppet::Interface.new :name do ... end
There are impli... -
- Interface declaration files (lib/puppet/interface/<interface_name>.rb) and action declaration files (lib/puppet...
-
- I believe this ticket will be rendered unnecessary by ticket #6781. Some of Dan's logic may need to be moved from th...
-
- The `--format [TYPE]` flag should be usable for both serializing output data and deserializing data (primarily from S...
-
- Tests were written in 2.6.x that specify a behavior of mount refresh that is different than with this patch. I'm not...
-
- The expectation is that actions should be able to transparently have typechecked signatures, and that automatic coerc...
-
- When I try to run the unit tests after installing from source I get the following error:...
-
- When I try to run the rspec tests after installing from source I get the following error:...
-
- I tried using C:\Tools\puppet. This ran without a failure message. It created the dir and one file in that dir: C:\...
-
07:54 pm Refactor #6776 (Duplicate): MS windows should have it's own exec provider
- Now that the exec type groks providers, having the posix provider support both MS Windows and *nix is needlessly comp...
-
- I don't want to assume what this directory is for. I would guess this is where it is going to execute out of, be ins...
-
07:46 pm Bug #6775 (Closed): puppet master --genconfig bug
- looks like --confdir and --vardir is inconsistent and other variables miss declared default
$ puppet master --... -
- The workaround for this would be to specify the directory on the command line; e.g. something like:
C:\Project... -
- One useful convention established by other command-line scripts is to read arguments from ARGV by default, opting to ...
-
- When I run the install.rb script from source I get the following failure message:...
-
- I've been trying to find out which manifest file is causing this, but I have quite a few that I just updated.
Both... -
- Setting the status to "Investigating" was accidental. Thought I had set it to "Accepted."
-
- Some of us (but not others) have been getting a mysterious failure when running the spec tests:
Failures:
... -
- In order to provide for better backward compatibility if a specific interface changes in the future, we should have a...
-
- We've had several requests for a self-service capability in the Dashboard. The user stories look like:
"We have d...
03/18/2011
-
03:57 pm Feature #5528: API call to sign, clean or generate certificate request
- How does your work mesh with this: [https://projects.puppetlabs.com/projects/puppet/wiki/Certificate_Authority_Rest_...
-
03:03 pm Bug #6767 (Duplicate): External nodes ignore missing Classes
- Puppet completely ignores missing classes for external nodes as you can see here:
http://groups.google.com/group/p... -
01:53 pm Feature #6766 (Accepted): Ability to display messages during agent run without creating resource ...
- The problem with the notify {} resource is that it isn't idempotent, and always shows up as a change.
Much like no... -
01:33 pm Facter Bug #4407: Fact names should be sanitized to be alphanumeric
- I am not able to reproduce this issue, and it seems to have been fixed a long time ago, in commit aa56886b1143d49173f...
-
- Alright, I've got a patch uploaded and sent to the list for review.
-
- I have a fix for this (using Facter::Util::Resolution.exec). As soon as I get the git server I've been using fixed u...
-
11:23 am Facter Bug #6763 (Closed): arp uses incorrect path for RHEL/CentOS installations
- Facter 1.5.9 looks for /usr/sbin/arp, but RHEL5.6 net-tools installs it on /sbin/arp
[root@test ~]# facter
... -
10:31 am Feature #5824: Puppet language should have self keyword
- This would actually make a lot of sense. Certainly it would clarify the code and simplify this form of expression.
-
-
- I'm just going to say that undef is sufficient. In cases where it is not, it is possible to work around it with the e...
-
- This also makes adding plugin actions to an existing interface impossible as well.
Ex:
I would like to add an a... -
- Please excuse my ignorance with SUSE packaging here, but is there a reason to include the two patch files in /usr/src...
-
- Luke Kanies wrote:
> We already use gnu style options, so that wouldn't be a behavior change at all.
>
> You're r... -
- I can't reproduce this. More specifically, I can't reproduce the *success* case reported: everything from 2.6.0 thro...
-
- We already use gnu style options, so that wouldn't be a behavior change at all.
You're right that this would be in... -
- Nigel Kersten wrote:
> I agree.
>
> We should just drop array support for all singleton value'd attributes. It do... -
- Luke and I have done some initial work on prototyping an implementation of this, which we (well, he) expect to flush ...
-
- I would strongly encourage us to head down the path of GNU style (and, er, {Ruby, Perl, Python}-style) option handlin...
-
- This looks like a net loss to me: we have to open-code all the option handling. Can't we delegate this to one of the...
-
12:57 am Facter Bug #6515: processorcount not initialized
- Added fixtures and tests for cpuinfo for amd64 machines (see github)
-
- In the current implementation of interfaces, when invoking an indirector action through the command-line you may spec...
-
- The code base contains many occurrences of the pattern:...
-
12:18 am Facter Bug #6740: facter doesn't always respect facts in environment variables
- The initial patch was intended to be low impact, but here's a more generic version. It renames the length method to ...
-
- An additional problem with the existing mechanism is that if a user were ever to try to create an interface action ca...
-
- Digant, are you actually collecting resources?
Have you observed reads occurring when you didn't expect it?
Do you ... -
- In the current implementation of interfaces, in order to pass an option to an interface action, one must set the opti...
-
- Integrated second patch and added some unit tests using these cpuinfo fixtures.
-
- Dom, any chance you can attach a "puppet --genconfig" to this so we can see what the report related settings are all ...
-
-
-
- Available in master in commit:2b8be23d1b670a8664fb5aa74134b0786c0a9fa7.
This adds a --plugin REPO option. -
- Installing plugins from git is necessary primarily to test that puppet still functions correctly with puppet-interfac...
-
- The plugin form of interfaces requires a small file for each indirection noun of the form:...
-
- The current implementation of interfaces allows a render method to be specified at the interface level but not at the...
-
- Are you thinking of #5794 ?
-
- Slightly modified version merged as of commit:330d3992744b80187223e5b6af2a58219e4610f6
-
- I swear we have another bug around about this... but I thought it had been merged into 2.6.next by now.
Investigat... -
- It's unclear to this point what this means in terms of backwards compatibility. Currently, with existing Puppet appl...
-
10:48 pm Bug #6751 (Duplicate): Reports directory not created on Agents
- This defacts causes System Acceptance test failures on hudson running againt 2.6.next:
version facter=>"1.5.8-93-g... -
- API design is part of this ticket. Don't necessarily just take the above description as the desired implementation.
... -
- Posix style, in this context, means that Puppet needs to distinguish between options specified in three possible loca...
-
- In the current implementation of interfaces, options may only be specified at the interface level. This means that t...
-
10:13 pm Bug #6748: ndefined method `<<' for {}:Hash when combining resource chaining, metaparameters, and...
- I found the issue.
Using the -> relationship syntax should work between this user and group resource, but does not.
... -
- # Overview
Getting a difficult to diagnose error. No idea what's failing in the parsing of the manifest since I d... -
- Daniel reviewed the final changes today, and I just merged the branch in in commit:462a560
-
08:21 pm Bug #6310: md5sum reports different md5 than state.yaml
- Matt Robinson wrote:
> Could you provide us the contents of a file where you see this behavior? That would certainl... -
- I agree.
We should just drop array support for all singleton value'd attributes. It doesn't make sense to support ... -
- Currently the timeout parameter to the exec type will accept an array as input, but the tries and try_sleep parameter...
-
- James Turnbull wrote:
> Seems reasonable.
Thinking more about this, I am getting less and less convinced: I can't... -
- I didn't. I'm about to now.
-
- Yep. We're still going to follow the process in update 20.
-
- Hey. Thanks for making this contribution. I took a look, and it doesn't look wrong, but...
Specifically, the "bu... -
- Thanks for the fixes and comments. Revised accordingly.
-
- Just so people have history on this:
Puppet works this way because it was written to be compatible with cfengine, ...
03/17/2011
-
04:21 pm Bug #6745 (Duplicate): Puppet needs a better undef/way to handle nil's from custom functions
- It appears to me that undef isn't really a real type in puppet. If I pass something that is undef to a custom functi...
-
- If you assume people are taking advantage of every case where we don't raise an error and should, then we'd never end...
-
- We are writing several routes like this:
define route::multiple( $routes) {
file { "routes":
name => $o... -
- This is a very serious issue for me. Postgresql databases on Debian are initialized to the locale used to install it,...
-
- Daniel Pittman wrote:
> Nigel Kersten wrote:
> > What's blocking on my input here?
>
> Damn. Nothing. I though... -
-
-
- Seems reasonable.
-
-
- This is a 1.5.9 regression.
-
- Duplicate of #4980 and fixed there.
-
-
- I slightly mispoke in my previous entry:
"there are people"
should have been
"there must be people"
In g... -
10:13 am Facter Bug #6728: Facter improperly detects openvzve on CloudLinux systems
- Ben Hughes wrote:
> Is this for the virtual fact?
is_virtual improperly set to "true", must be "false"
virtual i... -
- Thank you for updating this. I'll build a SUSE VM and get this added.
-
- Ah cool. Thank you for letting us know. Passenger/apache scales much better anyway.
-
- Thanks, that branch tested fine for me.
-
- Have a nose at https://github.com/barn/facter/tree/tickets%2Fmaster%2F6716 and see if that works for you please?
-
- I have just test 1.5.9rc1 and found that I have a similar issue, but with ipv4 only.
dhcp-au-68:facter mike$ R... -
- I hate bikeshedding these things up, but this should really be uniform across the entire suite of network facts that ...
-
- https://github.com/puppetlabs/puppet/commit/489942706726629fe0d477ebb2692c2b0b05c43e
Now available in `next`; we m... -
- Nigel Kersten wrote:
> What's blocking on my input here?
Damn. Nothing. I thought I changed that field to me. ... -
- Sounds reasonable.
I'm passing over to James for comment, as he's more aware of the areas of the website this may ... -
- https://github.com/richardc/facter/tree/ticket/master/6740
-
- What's blocking on my input here?
-
- It will be targeted at Statler, the codename for our next non-maintenance release.
-
- After chatting to some people, it seems like the current behavior is relied upon.
The current approach is:
* ... -
- On an OSX host:
$ facter operatingsystem
Darwin
$ facter_operatingsystem=Not_Darwin facter operati... -
- Dan Bode wrote:
> Then I reascend my efforts as I cannot support that decision. As crazy as it sounds, there are peo... -
- We have CHANGELOG and RELEASE_NOTES.md which are essentially the same info twice. The RELEASE_NOTES.md is linked to ...
-
- Given it more thought, dangerous I know.
Extreme fringe case alert.
1) Would cause problems for IPv6 only hosts... -
- Is this for the virtual fact?
Is CloudLinux not just a distro running atop of openvz? -
- https://github.com/daniel-pittman/puppet/tree/feature/next/6722-it-should-be-possible-to-run-parser-function-rspec-te...
-
09:28 pm Bug #6738: Mount resource in puppet duplicates ramdrive mounts for each puppet run
- code fix:
$size = "128"
$ramdisk_size = "size=${size}m"
-
- If ensure => mounted is used the ramdisk is mounted multiple times (once every puppet run), even though only one entr...
-
09:20 pm Bug #5876: Require and Subscribe on the same refreshonly exec doesnt work
- Just to chime in, also affecting me on 2.6.6:...
-
- I am not sure if this is related - i can make a new bug if you think it isn't - but I am seeing similar notify issues...
-
- Can we update the status if it's not just "Accepted" anymore?
-
- Yes - which sucks but ... :(
-
- Thanks. Er, I am a little concerned that we take a single, static snapshot of used memory, then return that forever,...
-
- commit:d62e079489c07201cb343f2ca109fecd62d6e567
-
- Then I reascend my efforts as I cannot support that decision. As crazy as it sounds, there are people in the wild tak...
-
- Yep - just had not hit enter.
-
- I refactored the provided code a little to make it easier to test and added a few tests, but the tests currently fail...
-
- Dan, we are targeting the complete fix at 2.6.x, which means failing just like we do when you include a class from ma...
-
- Removed erroneous branch location.
-
- The puppet master expects, if you are using the system-wide paths, to be invoked as root. It will then internally dr...
-
- Implementing the feature as discussed in this ticket would be too much of a functionality change to be considered for...
-
- Two bits of feedback, one relevant and one irrelevant:
The relevant one is that your commit doesn't implement what... -
- I can't find this in the commit mentioned, or the ticket, in the upstream repository testing branch?
-
- I can't see that commit in the upstream repository. Are you sure that is the correct hash?
-
- Checked in as commit:284ecb7fa2a21c85b617f2b9c21bdff661c89001.
Added `preinitializer.rb`, which is automatically r... -
- commit:e917e1a7f47b4399f5d1dd32dd5cd8ad5da86b64
-
- I am going to take this one b/c its so easy and I am sick of being bitten by it. I am hoping to get it into 2.6.x so ...
-
-
- Now pushed into master, and available everywhere.
-
- commit:d0f0f63f91a0eff72c1c528a6b5fb325bd9001d4
-
- Against recent versions of RubyGems, there's an error in `vendor/rails/activesupport/lib/active_support/dependencies....
-
- Nigel, did you send out that proposal? I'm updating the release notes for 2.6.7rc1 and was trying to figure out what...
-
- This problem breaks hudson testing for 2.6.next
version facter=>"1.5.8-93-g469d2a2", puppet=>"2.6.4-493-gea34876"
... -
- At the moment we have embedded the ESX password in our public repository, which is pretty much suboptimal.
After d... -
- Merged upstream.
-
06:04 pm Feature #6607: Add support to control verbosity of masterhttp.log
- That's great, thanks for your help!
Which release are you expecting this feature to be added?
Cheers,
Armando
-
- Now merged in commit:c60c6cacaca6b8e34c29835f9e5749fc380b8e0b
03/16/2011
-
-
-
- commit:139010043660a51287f6b913e6960dd65d3a3e9c
-
- commit:1207765dde99a84ca49a99058c66db7bfa5ca0dd
-
- commit:ffd80ac4e48285121e6557a06172d47d86270a76
-
- Facter 1.5.8 uses following algorithm to diagnose OpenVZ VE:
def self.openvz?
FileTest.directory?("/pro... -
- Stefan's patch fixes the problem and has no side effects I can find. I updated the patch with some more extensive te...
-
- I'm happy to reopen it, but if someone could say "yes this still a problem in 2.6.x" then that would be great.
-
-
- Should or does it distinguish itself in some way from say, other Solaris/OpenSolaris systems?
-
- I am a little confused about this ticket being closed. It is still a reasonable feature request and has not been impl...
-
- Closed due to inactivity.
-
- Closed due to inactivity.
-
- While not a fix as you ask for. You can add keylength = however-big-you-want in the "main" stanza in your puppet.conf...
-
-
02:40 am Bug #6727 (Closed): "Could not find user" When Managing a File - But the New User Exists
- A basic user and file creation like this
$user_name = 'puppet_test'
$group_name = 'puppet_test'
@g... -
12:51 am Bug #5610 (Merged - Pending Release): hasstatus breaks dependency of named service resource.
- Available in 2.6.next as of commit:4cb2f01884e04a8e08fb4eee18232412d4194b6c
-
-
- OK: I'm not satisfied that I can get this working before I leave on vacation.
I really wanted to merge your submit...
Also available in: Atom