Perl 6 - the future is here, just unevenly distributed

IRC log for #puppet-openstack, 2014-02-17

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:17 kitp joined #puppet-openstack
00:17 ryanycoleman joined #puppet-openstack
01:15 openstackgerrit Sebastien Badia proposed a change to stackforge/puppet-cinder: multi rbd: add opts, deprecate glance_api_version  https://review.openstack.org/73892
01:18 kitp joined #puppet-openstack
01:27 fandi joined #puppet-openstack
01:35 kkirkpatrick joined #puppet-openstack
02:17 xarses joined #puppet-openstack
02:18 kitp joined #puppet-openstack
02:20 xingchao joined #puppet-openstack
02:34 fandi joined #puppet-openstack
03:02 ianw joined #puppet-openstack
03:29 otherwiseguy joined #puppet-openstack
03:34 fandi joined #puppet-openstack
03:57 kkirkpatrick joined #puppet-openstack
04:38 openstackgerrit A change was merged to stackforge/puppet-openstack: Remove deprecated address from keystone spec  https://review.openstack.org/71990
05:18 kitp joined #puppet-openstack
05:34 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-keystone: Load tenant un-lazily if needed  https://review.openstack.org/73913
06:15 saju_m joined #puppet-openstack
06:19 kitp joined #puppet-openstack
07:20 kitp joined #puppet-openstack
07:27 dgollub joined #puppet-openstack
07:32 openstackgerrit A change was merged to stackforge/puppet_openstack_builder: Propagate 'verbose' and 'debug' to the nova module  https://review.openstack.org/73626
07:59 EmilienM good morning
08:21 kitp joined #puppet-openstack
08:22 bauzas joined #puppet-openstack
08:33 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-keystone: keystone-client: update endpoint argument  https://review.openstack.org/73970
08:38 michchap EmilienM: morning
08:39 EmilienM michchap: bonjour !
08:39 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-keystone: keystone-client: update endpoint argument  https://review.openstack.org/73970
08:41 dgollub joined #puppet-openstack
08:43 mattymo ahoy, gentlemen
08:43 michchap mattymo: morning
08:44 mattymo for simple MySQL deployments (non-HA), what custom my.cnf options do you use?
08:44 michchap while both of you are here: I've got local patches bumping the mysql module up from 0.x to 2.2, but it's pretty disruptive. Thoughts on a timeline for bumping the version for all the modules?
08:44 mattymo I know galera has its own set of optimizations and I've already gathered lots
08:44 mattymo michchap, what are you changing with mysql?
08:45 michchap the puppetlabs-mysql version, from 0.x to the latest master
08:45 michchap it's been rewritten and it's a ton cleaner now. The types are all changed and some of the classes have been moved around
08:48 mattymo wow it is a lot better
08:48 michchap I'm considering adding a mysql_module_compat parameter and just adding support for the new version right now, but migration will take time
08:48 michchap yeah it's sweet. I'm using it for galera :)
08:48 openstackgerrit A change was merged to stackforge/puppet_openstack_builder: Remove unused file  https://review.openstack.org/73499
08:49 michchap because of the strict ordering down the bottom of mysql:;server, you can 'inject' steps in the install very easily, so getting it to manage galera wasn't too hard.
08:49 mattymo so actually we're still moving from 2.7 to 3.x
08:49 michchap re: conf options, I have almost no wisdom to share sadly
08:49 mattymo and I told my team last week that strictordering exists in 3.x. they almost fainted
08:49 michchap puppet version?
08:49 mattymo yeah
08:49 michchap ah
08:49 michchap --ordering manifest
08:49 mattymo yeah
08:50 michchap it's like you're running chef!
08:50 mattymo I still never ran chef
08:50 michchap I've not used it for anything either
08:51 michchap mattymo: is there any mirantis material on configuring mysql correctly for different cluster sizes?
08:51 mattymo we only have 2 options
08:51 mattymo single cluster with mysql and HA with galera on 3 controllers
08:52 michchap you don't scale past 3 because something starts to slow down...rabbit or mysql?
08:52 mattymo we don't scale past 3 because the load on controllers shouldn't be so high
08:52 mattymo you can manage 200 computes on 1 powerful controller
08:52 michchap ah thanks
08:53 mattymo and if you need more than 200 compute nodes, you probably want a special architecture anyway
08:53 michchap I saw https://tickets.opscode.com/browse/COOK-2100 linked on the ML
08:54 mattymo yes you must rebuild your log files in order to change that parameter
08:54 mattymo or start over
08:54 mattymo 5mb logfile default doesn't affect performance enough to make a difference
08:55 michchap I did wonder whether it made much difference. 5mb is an enormous amount of log data
08:55 mattymo log is mysql transaction log
08:55 mattymo it will grow over 5gb on a prod cluster
08:55 mattymo it's okay
08:55 michchap but over what period?
08:56 mattymo depends on how many users and how hard you hit it. If you use rally to punish your cluster, you can do that in a day
08:56 openstackgerrit Sebastien Badia proposed a change to stackforge/puppet-cinder: multi rbd: add opts, deprecate glance_api_version  https://review.openstack.org/73892
08:56 mattymo we're considering options to prune it
08:56 mattymo I don't know if you can
08:57 michchap does it need it so that it can replay transactions?
08:58 mattymo yes
08:58 mattymo so simple mysql replication (non-galera) relies on that for sure
08:58 saju_m joined #puppet-openstack
08:58 mattymo but I don't know if galera needs it
08:58 michchap ah. I'm really not experienced at all as an sql admin
08:59 michchap I just hoover up guides and convert them to puppet
09:00 mattymo ha same here
09:14 derekh joined #puppet-openstack
09:15 starmer joined #puppet-openstack
09:18 dgollub joined #puppet-openstack
09:20 mmagr joined #puppet-openstack
09:21 kitp joined #puppet-openstack
09:49 openstackgerrit Benedikt Trefzer proposed a change to stackforge/puppet-cinder: rbd: add params and deprecate glance_api_version  https://review.openstack.org/70229
09:53 bogdando joined #puppet-openstack
10:22 kitp joined #puppet-openstack
10:41 ryanycoleman joined #puppet-openstack
10:48 ryanycol_ joined #puppet-openstack
10:55 ryanycoleman joined #puppet-openstack
11:23 kitp joined #puppet-openstack
12:24 kitp joined #puppet-openstack
12:57 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: attach disks when running integration tests  https://review.openstack.org/74028
13:01 markvoelker joined #puppet-openstack
13:08 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: attach disks when running integration tests  https://review.openstack.org/74028
13:08 openstackgerrit A change was merged to stackforge/puppet_openstack_builder: Expose syslog parameters: use_syslog and log_facility  https://review.openstack.org/73801
13:14 openstackgerrit Pierre Rambaud proposed a change to stackforge/puppet-nova: Fix usage of double quote instead of simple quote  https://review.openstack.org/74033
13:22 openstackgerrit Pierre Rambaud proposed a change to stackforge/puppet-nova: Fix usage of double quote instead of simple quote  https://review.openstack.org/74034
13:24 openstackgerrit Pierre Rambaud proposed a change to stackforge/puppet-nova: Fix usage of double quote instead of simple quote  https://review.openstack.org/74036
13:25 kitp joined #puppet-openstack
13:28 openstackgerrit Loic Dachary proposed a change to stackforge/puppet-ceph: add the mon string in class names  https://review.openstack.org/74038
13:33 rcrit joined #puppet-openstack
13:34 rcrit left #puppet-openstack
13:43 dprince joined #puppet-openstack
13:48 Butcho joined #puppet-openstack
14:01 prad_ joined #puppet-openstack
14:04 Butcho joined #puppet-openstack
14:07 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:09 michchap mattymo: ^ take a look at that. I've tried not to be disruptive, and the cinder module is one of the cleaner ones, but that mysql_module param is going to have to be added to everything that touches a db class, including allt he ones that include or require the mysql::python bindings, which is now under mysql::bindings::python
14:10 michchap mattymo: but I might be wasting my time...I'm guessing bumping the mysql module version in Fuel or RDO or whatever is a lot of work.
14:12 mattymo why swedish?
14:13 openstackgerrit A change was merged to stackforge/puppet_openstack_builder: Added ipmi installation to the Cisco vendor script.  https://review.openstack.org/73654
14:13 michchap it's the mysql default for some stupid reason
14:13 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:13 mattymo that sucks
14:14 michchap we can change the default charset to utf and be more sensible, but that's for another day
14:14 mattymo yeah we're using utf8 in fuel
14:15 mattymo but it looks ok to me
14:15 michchap mattymo: makes sense
14:16 mattymo I really think a master openstack module would make db management better
14:16 michchap mattymo: I discovered exactly that while doing this
14:16 mattymo since all modules implement sqlalchemy relatively the same
14:16 mattymo we want to revive puppet-openstack and put logging and db in there correctly
14:16 michchap yes
14:16 michchap and firewall
14:16 mattymo we have our own insane firewall module
14:17 michchap oh ok, well I'll put one in that just does the basics with puppetlabs-firewall
14:18 michchap there's a message on the list from chris hoge in the thread from mgagne about keystone auth, and in it chris proposes openstacklib or similar
14:18 kitp joined #puppet-openstack
14:18 mattymo oh we are using puppetlabs-firewall too
14:18 michchap I think a clean slate is probably needed.
14:19 mattymo let me see if our patch made it upstream for evaluating cidr
14:20 mattymo oh no that was something we wrote bad on our own
14:20 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:24 kitp joined #puppet-openstack
14:29 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:33 Butcho joined #puppet-openstack
14:34 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:37 dachary morning puppet
14:45 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-glance: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74061
14:46 michchap morning dachary :)
14:48 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-cinder: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74048
14:50 openstackgerrit Michael Chapman proposed a change to stackforge/puppet-glance: Add support for puppetlabs-mysql 2.2  https://review.openstack.org/74061
15:04 otherwiseguy joined #puppet-openstack
15:09 rwsu joined #puppet-openstack
15:14 dmsimard joined #puppet-openstack
15:16 dmsimard dachary: Looks like puppetlabs just killed puppet-rspec-system: https://github.com/puppetlabs/rspec-system-puppet
15:16 ken_barber dmsimard: actually its superseded by beaker
15:16 openstackgerrit Pradeep Kilambi proposed a change to stackforge/puppet-ceilometer: Fix alarm package params for ubuntu havana  https://review.openstack.org/71679
15:16 ken_barber dmsimard: but yeah, basically I've killed it
15:17 dmsimard ken_barber: Got to check the features but there was this one issue that was just closed that I was looking forward to :(
15:17 dmsimard ken_barber: "Add a way to parse a Modulefile and install dependencies for a module"
15:17 ken_barber dmsimard: yeah, you're going to need to move that to beaker I guess. sorry :-/.
15:21 mattymo that still doesn't exist?
15:26 dmsimard mattymo: I don't know, I haven't followed the project in a bit and pretty much haven't looked at beaker at all yet
15:26 dmsimard First time I hear about beaker
15:30 mattymo oh beaker me too
15:30 mattymo we still don't use per-module puppet testing in Fuel
15:31 mattymo just complete integration tests that take ~45 mins
15:48 mjblack joined #puppet-openstack
15:49 mjblack joined #puppet-openstack
15:52 dmsimard dachary: Why not host the fork on github so we can keep track of various forks (repository network) ? If puppetlabs is deprecating rspec-system and we want to keep using it, we can become some sort of new upstream.. ?
15:54 ken_barber dmsimard: I'd honestly just switch. beaker-rspec is designed to be a close match to rspec-system, and since most of the core modules are switching you'll just be special casing the openstack stuff. Of course, you can do whatever you prefer, just my 2p.
15:56 dmsimard ken_barber: I'm not very familiar with rspec-system, dachary is the one that worked on the implementation for puppet-ceph. Can we get an actual README on beaker so we know what it is about? :)
15:58 ken_barber dmsimard: so right now there is only the core project that is well documented: https://github.com/puppetlabs/beaker ... the rspec bridge docs are well, poor as you know: https://github.com/puppetlabs/beaker-rspec. I believe ashp and the others are working towards a reasonable readme.
15:58 ken_barber dmsimard: I think for now they are just pointing people at the working 'examples' in puppetlabs/apache, puppetlabs/postgresql and the other main core ones.
15:58 ken_barber dmsimard: which I agree, is less then desirable.
16:00 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-nova: compute: RBD backend support  https://review.openstack.org/72440
16:03 mgagne joined #puppet-openstack
16:10 fvollero EmilienM: https://review.openstack.org/#/c/73620/2 what you think about this ?
16:10 EmilienM fvollero: no idea
16:10 EmilienM yet
16:11 fvollero EmilienM: i was thinking about it on 14 but honestly, I need to test it
16:12 kitp joined #puppet-openstack
16:13 dmsimard EmilienM: I'd appreciate if you could look at https://review.openstack.org/#/c/73650/ - there is a review for the backport to havana as well
16:13 EmilienM dmsimard: I'll have a look asap, ok
16:13 dmsimard EmilienM: <3
16:21 openstackgerrit A change was merged to stackforge/puppet-swift: Add support for middlewares with hyphens  https://review.openstack.org/72693
16:23 openstackgerrit David Moreau Simard proposed a change to stackforge/puppet-swift: Add support for middlewares with hyphens  https://review.openstack.org/74082
16:23 kkirkpatrick joined #puppet-openstack
16:23 dmsimard ^ havana backport
16:30 pasquier-s joined #puppet-openstack
16:33 dtalton joined #puppet-openstack
16:51 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-horizon: local_settings: Allow to manage can_set_password option  https://review.openstack.org/68433
16:54 EmilienM dmsimard: https://review.openstack.org/#/c/73650 is very good, but I think we could also set log_file
16:55 dmsimard EmilienM: true, believe it would be another commit, though.
16:55 EmilienM dmsimard: no I think it comes together, we have the same issue here
16:55 EmilienM i may be wrong
16:56 dmsimard Don't have my CI openstack cluster up right now or I'd check
16:56 dmsimard In the context of my usage, I think it ends up with the name proxy-server.log in /var/log/ceilometer
16:56 EmilienM ok
16:57 dmsimard I'll have my cluster up soonish
16:58 EmilienM dmsimard: cool, I have to go, I'll be back later
16:59 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-horizon: local_settings: Allow to manage can_set_password option  https://review.openstack.org/68433
17:07 markvoelker EmilienM: perhaps we can finally get https://review.openstack.org/#/c/70054 and https://review.openstack.org/#/c/68433 dealt with today. =)
17:11 bauzas joined #puppet-openstack
17:24 ianw joined #puppet-openstack
17:29 mgagne markvoelker: If find it "sad" that we have to use @merged_ variables, it's ugly.
17:30 markvoelker mgagne: Well, if puppet had mutability... =)
17:30 mgagne markvoelker: this syntax could be used to define default values: @variable ||= "default value"
17:30 mgagne markvoelker: but default values logic would then be moved to template
17:30 markvoelker ...which is also a bit ugly, no?
17:31 mgagne markvoelker: variable name would now be much "beautiful" =)
17:31 mgagne markvoelker: we are already using this syntax for keystone variables
17:32 mgagne markvoelker: or we can skip the value altogether if not defined, horizon should already have a default value anyway.
17:32 markvoelker mgagne: I'd prefer to have them in there.  Seems to make debugging easier for end users who aren't familiar with horizon's built-in defaults.
17:33 mgagne markvoelker: I could argue that there is probably much more than what's in local_settings.py
17:41 marun joined #puppet-openstack
17:43 markvoelker mgagne: Sorry, had to step away for a phone call.  Yes, there probably is more, but what's there seems to help users I've worked with out.
17:43 mgagne markvoelker: we should come up with a better name then @merged_ =)
17:44 mgagne markvoelker: as it's would now be part of the "interface" of our module and can't be changed easily later on.
17:45 markvoelker mgagne: well, if we're already shifting the default logic to the template for other things I'm not opposed to being consistent (I'm sure I really like putting default logic in templates vs classes, but that ship has apparently sailed).
17:46 markvoelker s/sure/not sure/
17:48 markvoelker The erb code will look somewhat different I suppose since this is a hash, but functionally the same.
17:57 mgagne markvoelker: Could merge with defaults be moved to template instead and neutron_options overridden there before it's used?
17:58 prad_ joined #puppet-openstack
17:58 dmsimard puppet-swift question, any ideas why these would be defines rather than classes ? https://github.com/stackforge/puppet-swift/tree/master/manifests/storage/filter  — They're analogous to proxy pipeline middlewares which are classes: https://github.com/stackforge/puppet-swift/tree/master/manifests/proxy
17:58 markvoelker mgagne: that's more or less what I'm looking at right now
17:59 dmsimard I would change them to classes but that pretty much breaks backwards compatibility.
18:01 mgagne markvoelker: not being able to reassign a variable is a PITA. Using intermediate variable looks to be the design pattern used in Puppet to bypass this limitation. It's ugly but yha, we have no choice in that regard.
18:01 mgagne markvoelker: However as soon this intermediate variable "leaks" to template, I'm concerned. Parameter is no longer a 1:1 to template and end user wil have to read the manifest to understand what's going on.
18:01 dmsimard Oh, I kind of get why they are set as defines ..
18:04 mgagne dmsimard: filter is shared between account, container and object.
18:04 dmsimard mgagne: Yeah, that's what I realized from the usage of ${name}
18:04 mgagne dmsimard: classes can't be declared twice
18:05 mgagne dmsimard: might be worth introducing validation against $name (account|container|object) so users don't use it to declare a recon filter for baboon for example.
18:06 mgagne dmsimard: unless there is some weird use cases I'm not aware of.
18:06 dmsimard (baboon|badger|gopher), right, I get it :D
18:33 dtalton i thought it was mushroom?
18:33 dmsimard badger badger badger badger badger mushroom mushroom
18:48 ryanycoleman joined #puppet-openstack
18:54 openstackgerrit joined #puppet-openstack
19:08 EmilienM markvoelker: yeah, would be awesome
19:08 EmilienM markvoelker: i did a rebase on your patch today
19:22 rharrison joined #puppet-openstack
19:44 otherwiseguy joined #puppet-openstack
19:49 openstackgerrit Mark T. Voelker proposed a change to stackforge/puppet-horizon: Parameterize OPENSTACK_NEUTRON_NETWORK settings  https://review.openstack.org/70054
19:51 markvoelker EmilienM: mgagne and I were chatting earlier.  He didn't care for the presence of @merged_ variables in the template, so I've just refactored a bit.
19:51 EmilienM markvoelker: just seen that, cool. I'll rebase tomorrow
19:51 mgagne s/didn't care/care/
19:58 kkirkpatrick joined #puppet-openstack
20:19 kkirkpatrick joined #puppet-openstack
20:25 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-nova: compute: RBD backend support  https://review.openstack.org/72440
20:31 openstackgerrit Emilien Macchi proposed a change to stackforge/puppet-nova: compute: RBD backend support  https://review.openstack.org/72440
20:32 Stefan1 joined #puppet-openstack
20:32 Stefan1 Anybody I can bug online?
20:33 Stefan1 I am having some issues getting the puppet ceph working :(
20:35 dmsimard1 joined #puppet-openstack
21:04 dtalton Stefan1, what's up?
21:15 ianw joined #puppet-openstack
21:19 Stefan1 oops
21:19 Stefan1 missed you
21:19 Stefan1 I had some issues with the ceph_config type
21:19 dtalton np
21:19 Stefan1 but  I fixed it - dummy me forgot to install the inifile module
21:20 Stefan1 now I stumped as the osd seemed to be described as resources on the git page
21:20 dtalton enovance's module?
21:20 Stefan1 https://github.com/stackforge/puppet-ceph/blob/master/USECASES.md
21:20 Stefan1 nope this one here
21:20 dtalton they use defined types for osd
21:21 Stefan1 but to me it seems to be regular classes
21:21 Stefan1 but classes are singletons - not?
21:21 dtalton oh yes, that module isn't finished
21:21 Stefan1 ah :-(
21:21 Stefan1 I guess I was too optimistic
21:21 dtalton so the ones that are are github.com/enovance/puppet-ceph (a very good modules)
21:21 Stefan1 is the enovance stuff any good
21:21 Stefan1 ah ok - got you
21:22 dtalton and github.com/dontalton/puppet-cephdeploy if you want to use cephdeploy to roll it out
21:22 dtalton enovance uses defined types for osds
21:22 dtalton puppet classes are singletons
21:23 Stefan1 your module is for openstack roll outs?
21:23 dtalton mine can be used standalone or with puppet_openstack_builder (the next-gen puppet-openstack)
21:24 Stefan1 okidoke
21:24 dtalton are you trying to use it with openstack?
21:24 Stefan1 I will check out the enovance for now
21:25 dtalton might be easier to just use ceph-deploy
21:25 Stefan1 Aaargh this is complicated :-)
21:25 dtalton heh
21:25 dtalton ceph-deploy is the official ceph tool to create clusters
21:25 dtalton http://ceph.com/docs/master/start/quick-ceph-deploy/
21:25 Stefan1 we a re working on an article series showing how to roll out a private cloud for continuous delivery
21:26 Stefan1 we set up a foreman at work to test everything with baremetal
21:26 Stefan1 now the collegaues responsible for the openstack
21:26 Stefan1 want to hand code everything
21:26 Stefan1 so bascially I am running after them and puppifying it all
21:28 Stefan1 enovance uses config files and "you" use the cli / admin tools?
21:28 Stefan1 ah nooo
21:29 Stefan1 I need to check the manuals in more detail
21:31 derekh joined #puppet-openstack
21:33 Stefan1 but thanks alot!
21:39 Stefan1 joined #puppet-openstack
21:48 starmer joined #puppet-openstack
22:00 sputnik13 joined #puppet-openstack
22:03 dtalton Stefan1, enovance does it the "old" way, which was to use system tools, mount disks manually, etc
22:03 dtalton puppet-cephdeploy wraps the new ceph cli deployment tool
22:06 dmsimard1 dtalton: It doesn't do it the old way, it does it the way it's meant to be - Alfredo (lead dev on ceph deploy) does not want people to wrap around ceph-deploy since ceph-deploy is meant to be used for testing and proof of concepts :)
22:06 dmsimard1 dtalton: He's not a fan of how Cisco and Mirantis (Fuel) are using ceph-deploy
22:06 dtalton dmsimard1, yes, I've heard :)
22:06 dmsimard1 dtalton: That doesn't mean it won't work, but it doesn't make the other method 'old' :p
22:07 dtalton dmsimard1, i misundestood the direction, i thought ceph-deploy was the end goal for all deployments
22:07 dtalton not just poc
22:08 dmsimard If I remember correctly, he said something along the lines of ceph-deploy takes certain shortcuts or does certain 'hacks' to get stuff done that you might not want/need in production - also about the limited latitude you might have regarding some features that ceph-deploy does not (or will not) support
23:10 starmer joined #puppet-openstack
23:48 ryanycoleman joined #puppet-openstack
23:48 JoeJulian Several neutron network options are not mutable once set but only a few are marked read-only. On the one hand, I feel that if a setting can only be set once, than it's not viable to be managed with puppet. On the other hand, I'd like to use puppet to set it up the first time so I'd rather not have any read-only attributes. Is there a philosophical consensus on this?
23:48 JoeJulian Frankly, I consider non-mutable openstack settings to be bugs but that's for a different channel...
23:50 rharrison joined #puppet-openstack

| Channels | #puppet-openstack index | Today | | Search | Google Search | Plain-Text | summary