Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2013-06-16

| Channels | #salt index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:01 the_drow Corey: http://pastebin.com/xa681CeH
00:01 Corey That's kind of a nutty state, but let's see here.
00:01 the_drow Corey: It's there!
00:02 the_drow Corey: and it's installing from the PPA.
00:02 Corey Comment:   Failed to confirm config of repo http://ppa.launchpad.net/fkrull/deadsnakes/ubuntu: repo "deb http://ppa.launchpad.net/fkrull/deadsnakes/ubuntu lucid main" was not found
00:02 the_drow so what gives?
00:03 the_drow Corey: I am using precise. The PPA is being added correctly. It's not the problem
00:03 Corey the_drow: You sure about that?
00:03 Corey cquinn@soap ~ % sudo add-apt-repository ppa:deadsnakes/ubuntu
00:03 Corey Error: can't find signing_key_fingerprint at https://launchpad.net/api/1.0/~deadsnakes/+archive/ubuntu
00:04 Corey If it doesn't work via manual calls, it won't work via salt. :-)
00:04 the_drow Corey: I just tested it.
00:04 the_drow Corey: works on my machine with manual calls
00:04 Corey Okay, let's test it on a precise node.
00:06 Corey the_drow: Worked there for me.
00:07 Corey State: - pkgrepo
00:07 Corey Name:      http://ppa.launchpad.net/fkrull/deadsnakes/ubuntu
00:07 Corey Function:  managed
00:07 Corey Result:    True
00:07 Corey Comment:   Configured package repo http://ppa.launchpad.net/fkrull/deadsnakes/ubuntu
00:07 Corey Changes:   repo: ppa:fkrull/deadsnakes
00:08 the_drow Corey: same. But the packages did not install
00:09 Corey the_drow: Are you sure they can coexist?
00:09 t0rrant joined #salt
00:09 the_drow Corey: yes. That's the whole point.
00:09 Corey Testing now.
00:10 Corey Hmm, *something* in my apt-get update is hanging and taking way too long, will have to track that down.
00:11 the_drow Corey: even if I write another package under pkgs: it still raises that exception
00:11 Corey [INFO    ] Executing command 'apt-get -q -y -o DPkg::Options::=--force-confold -o DPkg::Options::=--force-confdef   install python3.2 python2.6 python3.3 python3.1' in directory '/home/cquinn'
00:11 Corey Let's see if that melts. :-)
00:11 Corey the_drow: Link to the exception? Too many things floating around, don't want to get confused...
00:12 the_drow State: - pkg     Name:      python     Function:  latest         Result:    False         Comment:   An exception occurred in this state: Traceback (most recent call last):   File "/usr/lib/pymodules/python2.7/salt/state.py", line 1201, in call     *cdata['args'], **cdata['kwargs'])   File "/usr/lib/pymodules/python2.7/salt/states/pkg.py", line 516, in latest     cur = {desired_pkgs[0]: cur} KeyError: 0
00:13 pcarrier_ joined #salt
00:13 Corey the_drow: Your state just worked properly here.
00:14 davisamo joined #salt
00:14 the_drow Corey: https://gist.github.com/thedrow/5790143
00:14 Corey the_drow: And your master and minion are both 0.15.3?
00:14 fragamus joined #salt
00:14 the_drow Corey: no master.
00:15 Corey the_drow: http://pastebin.com/MffCzkwj
00:15 the_drow Corey: salt-minion is 0.15.3
00:15 the_drow Corey: Is your host 13.04?
00:16 Asphyxiated|Dk joined #salt
00:17 the_drow WTF??
00:18 Corey the_drow: Uh... you're on 13.04?
00:18 the_drow Corey: as a host. Yes.
00:18 Corey the_drow: ...and you're telling your PPA to install the precise version because...?
00:18 the_drow Corey: because the vm is 12.04...
00:19 Corey the_drow: Wait, what the hell are you doing? You don't have a master, you have a host and a VM...?
00:19 the_drow Corey: are you using Vagrant?
00:20 Corey the_drow: No, I didn't see that mentioned. That said, vagrant shouldn't change much either way-- it's just passing the states to the guest.
00:20 the_drow Corey: The laptop is the host for the VMs, Salt provisions the VM. The host is 13.04, the VM is 12.04
00:21 the_drow Corey: dafuq? sshing into the machine and running salt-call state.highstate -l debug made it work
00:22 the_drow Corey: and now that they are installed no exception is being raised.
00:24 the_drow Corey: can you explain that?
00:24 Corey http://t2.gstatic.com/images?q=tbn:ANd9GcQgLZGpP8AVNm9ImNjWLvaA4RzbpgavbzWxWZhF5XY7tfrMuuNsfg
00:25 the_drow Corey: So should I attempt to reproduce the error?
00:25 Corey the_drow: Sure thing. If you can, file a bug against salt-vagrant.
00:25 Corey I don't run it myself, so... :)
00:26 danieljn joined #salt
00:35 LarsN making progress.
00:36 LarsN ssh_interface: public_ips
00:36 LarsN helped with the auto-deleting instances.
00:38 the_drow Corey: dafuq again, I can't reproduce the error now :P
00:38 Corey LarsN: OOh.
00:38 Corey the_drow: Quinnteresting.
00:38 fragamus auser are you here
00:41 aat joined #salt
00:43 aat joined #salt
00:48 zonk1024 joined #salt
00:48 zonk1024 >.>
00:50 zonk1024 having issues with "salt '*' pip.freeze" saying it can't activate a virtual env. do I have to pass it something to tell it to do system-wide packages? it looks like it should handle it w/o any args
01:03 fragamus hey guys anyone using salt-cli?
01:07 druonysus joined #salt
01:07 druonysus joined #salt
01:11 the_drow Corey: how do I include file resources for masterless salt?
01:14 mgw joined #salt
01:16 fragamus argh im having trouble using salt-cli and i need this to work
01:35 mgw joined #salt
02:02 aat joined #salt
02:11 pcarrier_ joined #salt
02:24 davisamo I've got a quick question.  Libcloud is coughing up a lung when I try to do a salt-cloud --image-list aw2az1.      Could not find image 75845: <LibcloudError in None 'Could not find specified endpoint'>
02:24 davisamo I have these salt-cloud 0.8.8 installed
02:24 davisamo I have these two files defined in /etc/salt -- cloud.profiles  cloud.providers
02:26 davisamo aw2az1: identity_url: 'https://region-a.geo-1.identity.hpcloudsvc.com:35357/v2.0/' compute_name: Compute compute_region: 'az-1.region-a.geo-1' tenant: 'tenant' user: 'user' ssh_key_name: sshkeyname ssh_key_file: '/etc/salt/key.dat' password: 'mysupersecretpassword' provider: openstack script: bootstrap-salt-minion
02:26 davisamo and formatting is all screwy,  but that is the provider that I am trying to list all the images for
02:27 davisamo Can anyone give me an idea what I am doing wrong?
02:31 pcarrier_ joined #salt
02:49 fragamus argh im having trouble using salt-cli and i need this to work
02:51 zonk1024 joined #salt
02:53 fragamus https://gist.github.com/fragamus/5790569
02:53 fragamus MAYDAY MAYDAY MAYDAY
02:55 Corey fragamus: Calm down. :-)
02:57 Corey davisamo: provider: openstack  seems broken. :-)
02:58 fragamus MAYDAY :-)
02:59 Corey davisamo: Y'know, given that the rest of your stuff implies AWS.
02:59 Corey fragamus: Enough. Annoying people into supporting you is a poor strategy.
03:00 davisamo It isn't AWS.  It is hpcloud
03:00 davisamo and I did think openstack seemed rather broken.
03:01 Corey davisamo: Is that an openstack environment?
03:01 davisamo yes
03:01 Corey Provider is set to openstack, which is what makes me look at it oddly.
03:01 Corey (I don't have an openstack account to play with, unfortunately)
03:01 Corey Er, hp cloud rather.
03:01 davisamo no worries.  I thought I'd see if anyone noticed anything terrible about the config
03:01 davisamo Specifically,  it is a libcloud error
03:02 davisamo that I suppose I should be concerned about
03:02 davisamo I just have to trace the silly thing back to the source
03:02 Corey davisamo: " Could not find image 75845" is somewhat odd.
03:02 Corey davisamo: yeah, see if you can replicate it with curl.
03:02 Furao 400 Bad Request
03:02 Furao client request is bad
03:03 davisamo Yes,  I think that may have been an errant paste along with the error
03:03 davisamo You mean see if I can access the endpoint with curl?>
03:03 davisamo one moment
03:03 davisamo good idea...
03:04 Corey davisamo: Yeah, eliminate "the infrastructure" as a potential point of failure. :-)
03:05 davisamo pretty sure this is just user error... me <--
03:05 davisamo but checking anyway
03:05 Corey Yay. :-)
03:05 Corey Another satisfied customer!
03:06 davisamo Well,  if you could fix me maybe I'd be satisfied,  but you know how customers are... always unreasonable ;-)
03:06 davisamo lol
03:07 davisamo Thank you for taking a look at it.  I'll dig in again and see if I can figure out whats up
03:07 james4k joined #salt
03:07 Corey davisamo: No worries, let me know if you get stuck.
03:09 james4k hi, are there any open, full salt projects out there for reference?
03:10 james4k or maybe some thorough examples
03:10 EugeneKay james4k - https://github.com/saltstack/formulae
03:11 james4k cool, thanks EugeneKay
03:11 Corey The salt-states repo is old but good.
03:16 clintberry joined #salt
03:22 Furao james4k: https://github.com/bclermont/states/tree/master/states
03:22 Furao I think it's still the most complete set of example around
03:22 Furao one of the few file.accumulated usage
03:22 aat joined #salt
03:24 james4k thanks guys, these look great
03:30 LarsN where's the preferred location to file bugs?
03:31 Furao LarsN: github issue
03:31 LarsN thanks Furao
03:44 zonk1024 After Ansible Salt is so nice...
03:48 UtahDave joined #salt
03:48 Furao OMG I need to get one of these http://www.huffingtonpost.com/2013/05/07/poutine-soda-taste-test_n_3225268.html
03:49 zonk1024 joined #salt
03:59 Corey zonk1024: Why do you like Salt better than Ansible?
03:59 Corey zonk1024: I'm giving a talk on Salt soon, people are making noises about Ansible, it'd help to have a viable comparison.
04:00 UtahDave Corey: Have you tried the peer interface since Tom's latest fix?
04:00 Corey UtahDave: Not yet, have to upgrade it everywhere first. :-)
04:00 Furao Corey: wanna share your slides? I need to do one talk on a PUG here too
04:01 UtahDave coo
04:01 Corey Furao: I shared my slides ages ago from the last one I did, Ansible wasn't a thing yet.
04:01 Corey Redoing 'em from scratch this time, and it's not an "intro to salt" talk this time around.
04:01 LarsN UtahDave: how's it going.
04:01 LarsN have you done any work with salt-cloud on hpcloud recently(ish?)
04:01 Furao I wish there was some "open source" presentation of salt that can improved and evolve over time to ease propaganda
04:02 Furao a guy just did a talk on fabric :)
04:02 Corey Furao: Other people's material fits about as well as other people's shoes.
04:02 Furao yes, but it's easy to fork and change
04:03 Furao target the audience
04:03 Corey Furao: Sounds like a job for SaltStack's marketing folk.
04:03 LarsN davisamo: could you explain to UtahDave what it is you think you're seeing with the endpoints?
04:03 zonk1024 Corey: faster, register hosts, easy setup, modules look easier to create, nice syntax so far, faster, puppet-like state-based in addition to remote exe, faster
04:03 Furao http://www.justincarmony.com/slides/salt-intro/#/
04:04 UtahDave LarsN: hey, man!
04:04 Corey zonk1024: Seems like using Ansible to spark up Salt instances wouldn't be a horrid idea.
04:04 zonk1024 Corey: I would agree.  a basic bootstrap playbook would be nice
04:04 LarsN UtahDave: finally to the point where we're using Salt to get things rolling in my team.
04:05 UtahDave salt-cloud has a saltify driver now.  it installs salt on existing servers
04:05 LarsN Seem to have run into a problem, and it's entirely possible it's us doing things wrong.
04:05 UtahDave LarsN: Nice!
04:05 Corey That is an amazing presentation, Furao.
04:05 zonk1024 Corey: but for my ec2 stuff i'll roll it into an ami, so it'll try to connect when if spin it
04:05 Corey But I feel like it'd be outright theft to steal the jokes.
04:05 davisamo @UtahDave,  I put a couple of print statements in openstack.py and regardless of what provider I specify when doing a salt-cloud --image-list,  openstack.py is always getting the last provider's region.
04:05 LarsN UtahDave: davisamo has more information.  He's been spearheading the troubleshooting.
04:05 UtahDave What issue are you seeing, LarsN?
04:06 davisamo That makes me think that openstack.py is indeed only receiving the last provider listed
04:06 davisamo ... regardless of which one I specify
04:06 UtahDave hmm.  what version of salt-cloud are you on?
04:06 davisamo 0.8.8
04:06 LarsN on Ubuntu 12.04 iirc.
04:06 james4k how are deployments typically issued? using something like fabric?
04:07 james4k or am i missing something with salt itself
04:07 davisamo Acutally,  it is 12.10
04:07 UtahDave james4k: what do you mean? How do you install salt on your servers?
04:07 davisamo s/Acutally/Actually/g
04:07 LarsN james4k: if you're doing deployments into the cloud, salt-cloud can bootstrap the minion onto the instances after creation
04:07 LarsN which is what we're doing.
04:08 james4k deploying new app code to minions
04:08 UtahDave davisamo: have you tried using the latest from the salt-cloud git repo?  They're getting ready for the next release of salt-cloud. It's possible this bug has been fixed already
04:08 UtahDave james4k: Use Salt's remote execution to deploy code.  It's really fast and very powerful
04:09 davisamo No,  I haven't.  I can try that
04:09 davisamo FWIW,  it is failing on this call openstack.avail_images
04:09 davisamo in saltcloud/cloud.py
04:09 Corey james4k: While UtahDave is correct that Salt can easily deploy new code to minions, I tend to prefer to wrap said code in a deb or rpm first. :-)
04:10 james4k UtahDave, Corey: hmm ok, will look into it. happen to know of any good examples?
04:12 Corey james4k: What's your codebase written in?
04:12 zonk1024 Corey: another thing.  i just started playing with it today, and i've got syndic going through nat :D  salt is just fun.  (stepping out for a smoke)
04:12 Corey zonk1024: Nice! Haven't played with the syndic yet myself.
04:12 james4k Go, so in my case i am deploying a binary and assets it uses
04:12 Corey james4k: So it's what, a war file?
04:13 Corey james4k: Also, what distro?
04:13 james4k ubuntu
04:13 UtahDave LarsN: can I send you a pm?
04:13 james4k it is a tgz that atm i upload and unpack with fabric
04:13 LarsN sure
04:14 Corey james4k: I've written a war to .deb packaging stub that lives at https://github.com/kb1jwq/packaging-stub; if you take a look at the service stuff that lives in Salt's debian/ directory, I wrote a good deal of it.
04:14 Corey james4k: Unpacking a tarball is generally the job of a deb in this case. :-)
04:14 davisamo @UtahDave and @Corey,  the latest git seems to work fine.
04:15 Corey james4k: What I did in my last such gig was had Jenkins build a new package on every commit, I'd promote the debs I wanted to test into an equivalent deb...
04:15 davisamo question:  Why can't I do things the easy way?  :D
04:15 Corey Er, promote them to a proper repo, rather.
04:15 UtahDave davisamo: Ah, great to know.  There should be a proper salt-cloud release with the next salt release soon
04:15 Corey UtahDave: Oh good, more packages.
04:16 Corey UtahDave: Can't you just throw me down some stairs and call it good?
04:16 UtahDave james4k: Salt can push out your tgz and unpack it if you want.
04:16 UtahDave I love you, Corey!
04:16 * LarsN throws Corey down some stairs
04:16 LarsN at least it wasn't under a greyhound.  ;)
04:17 LarsN UtahDave: that sure was for you by the way,  you are welcome to PM me at will ;)
04:18 Corey james4k: UtahDave is of course correct, I'm just old and ornery and want to see things living in packages due to the benefits it gives you.
04:18 james4k i'll have to look into packaging into a deb, that sounds useful in itself
04:18 UtahDave yeah, that's always a good idea
04:19 Corey james4k: Feel free to ping me at any time, I had to struggle through it to learn how to package Salt, I'm better at it now. :-)
04:23 james4k thanks Corey
04:25 Corey james4k: No worries. Good luck!
04:25 erasmas joined #salt
04:38 james4k using remote execution for deployment, it seems it would need to be done from the salt master
04:39 james4k if you're deploying from a development machine to a production server, what does that look like?
04:41 UtahDave james4k: you can use the peer interface to allow one minion to send commands to another minion
04:43 james4k huh ok, so your dev machine would be a minion?
04:47 UtahDave sure, it could be
04:49 james4k i guess typically a build server would be
04:51 Corey james4k: Now you're thinking like a DevOps Unicorn!
04:52 Corey james4k: "Developers able to deploy code" doesn't mean they need root. :-)  Have the build server (Jenkins or similar) push code that passes tests.
04:54 james4k ah :)
04:55 davisamo left #salt
04:59 Corey james4k: And that push can be a manual "push the button" instead of a purely automatic process. There are degrees. :-)
05:02 elbaschid joined #salt
05:20 Furao joined #salt
05:36 sgviking left #salt
05:37 kmwhite joined #salt
06:25 erasmas joined #salt
06:27 kmwhite left #salt
06:31 kallek_ joined #salt
06:38 kallek joined #salt
06:46 james4k fyi, fpm seems pretty cool. makes deb creation simple and supports other package managers
06:48 yumike joined #salt
06:49 yumike joined #salt
06:49 yumike joined #salt
06:58 zz___number5__ joined #salt
06:58 Furao_ joined #salt
07:01 koolhead17 joined #salt
07:01 koolhead17 joined #salt
07:08 pentabular joined #salt
07:16 jpaetzel_ joined #salt
07:16 mackstic1 joined #salt
07:16 pt|Zool_ joined #salt
07:16 Happy86 joined #salt
07:17 kevinbrolly joined #salt
07:29 bhosmer joined #salt
07:29 faust joined #salt
07:35 sgviking joined #salt
07:46 linjan joined #salt
08:20 Nexpro joined #salt
08:27 elbaschid joined #salt
09:16 elbaschid hey guys, I've got a quick question about returners as I seem to have trouble to get the sentry returner to work
09:17 elbaschid is there a way to get a list of available returners?
09:18 azbarcea joined #salt
09:18 azbarcea_ joined #salt
09:19 zonk1024 elbaschid: https://github.com/saltstack/salt/tree/develop/salt/returners
09:23 elbaschid zonk1024: I've looked at that but don't seem to get it to work using 'sentry_return' or 'sentry' both cause an invalid returner message in the debug output of the minion
09:23 elbaschid is the correct way to specify the returner the module name or the name returned by __virtual__?
09:24 zonk1024 http://docs.saltstack.com/ref/returners/all/index.html#all-salt-returners   says sentry_return
09:24 zonk1024 =/
09:25 elbaschid Thanks for clearing that up :)
09:25 zonk1024 did you setup the connection? http://docs.saltstack.com/ref/returners/all/salt.returners.sentry_return.html#module-salt.returners.sentry_return
09:26 elbaschid yes, I've set it in the minion file where I use the returner
09:27 elbaschid also, when I run salt-master in debug mode it shows a message saying that sentry_return is loaded as virtual sentry
09:27 elbaschid I don't get any of that when running salt-minion in debug mode
09:28 zonk1024 guessing the returner is only server-side.  i'm not an expert on salt yet though.  give me a week
09:29 ronc joined #salt
09:29 elbaschid thanks for your input
09:29 elbaschid i'll play around with it a bit more and try to figure out what I am missing or doing wrong :)
09:36 zonk1024 elbaschid: firewall?
09:38 th3rt joined #salt
10:01 __gotcha joined #salt
10:01 __gotcha joined #salt
10:03 Furao elbaschid: sentry returner don't work in 0.15.3
10:04 Furao elbaschid: things changed in salt and the returner wasn't updated. here is a working version https://gist.github.com/bclermont/dc857c1c4abd8dd5c723
10:04 Furao and even that, in some cases it don't work, the bug in salt was fixed yesterday
10:05 Furao well that version in my gist use an other returner name to don't conflict with the one that come with salt
10:25 aranhoide joined #salt
10:26 EugeneKay Is there a saner/easier way to purge Pillar caches than `rm /var/cache/salt/minion/ -rf; service salt-minion restart` ?
10:46 santagada joined #salt
11:31 __gotcha joined #salt
11:31 __gotcha joined #salt
11:58 elbaschid Furao: thanks for that. You saved me a lot of time working this one out myself
12:00 elbaschid I might give your gist a try in the next few days and maybe look at some alternative returns too, although getting sentry to work would be ace :)
12:13 danieljn joined #salt
12:17 cossovich joined #salt
12:18 jslatts joined #salt
12:26 minioreo joined #salt
12:34 bhosmer joined #salt
12:34 Furao elbaschid: it's part my fault, I'm the author of this returner and I forgot to push back to saltstack the fixes
13:14 Newt[cz] joined #salt
13:18 fragamus joined #salt
13:19 cossovich I've got a state which runs an unzip command.... is there some way I can tell salt not to run the cmd if it's already performed it in previous run?
13:20 fragamus auser
13:20 Furao cossovich: https://github.com/bclermont/states/blob/master/states/_states/archive.py#L15
13:23 cossovich Furao: does archive work out of the box? or do I need to install that module?
13:26 jeddi joined #salt
13:28 Furao cossovich: it's a state I wrote, it don't depends on any module you need to follow these instruction to make it available to your minion: http://docs.saltstack.com/ref/states/writing.html
13:28 Furao and run saltutil.sync_all
13:28 Furao more here http://docs.saltstack.com/ref/states/writing.html#using-custom-state-modules
13:28 joehh Furao: what licence is that code under? Apache 2 like salt in general?
13:28 joehh the rabbitmq plugins looks particularly useful
13:29 joehh (and I think worthy of contribution to salt...)
13:30 cossovich Furao: got it. Thanks!
13:31 Furao joehh: these are outdated and might not work in salt 0.15
13:31 aranhoide joined #salt
13:31 Furao it's part of a product and service that I sell
13:32 Furao no website yet, as I already found some clients here and didn't had time to finish it
13:33 bhosmer joined #salt
13:36 timl0101 joined #salt
13:54 EugeneKay Who would I whine to about a deficiency in the salt-minion init script in the epel package?
13:55 EugeneKay It doesn't redirect 2>/dev/null properly, so I get lovely messages from yum after restarting the daemon and running anything that involves yum
13:57 EugeneKay (in a console on the minion)
14:00 __gotcha joined #salt
14:00 __gotcha joined #salt
14:03 th3rt joined #salt
14:03 __gotcha_ joined #salt
14:04 Kamal_ EugeneKay: a github issue would be the best
14:07 ronc_ joined #salt
14:10 aat joined #salt
14:16 EugeneKay Under saltstack/salt ?
14:26 aranhoide joined #salt
14:30 akoumjian joined #salt
14:31 akoumjian joined #salt
14:34 timl0101 joined #salt
14:47 efixit joined #salt
14:59 jeddi EugeneKay: if it's about salt-minion proper, then yes.
15:06 Newt[cz]1 joined #salt
15:19 ronc joined #salt
15:20 niki joined #salt
15:43 felskrone joined #salt
15:44 fragamus auser
16:06 clintberry joined #salt
16:20 Gifflen joined #salt
16:22 logix812 joined #salt
16:35 Corey EugeneKay: I've touched the logging for Salt itself, how sure are you that the bug lies in the init script?
16:36 EugeneKay Not very; I haven't tested it past noticing the yum messages on my shell
16:36 EugeneKay I'm going to fiddle with it some more before I file a report
16:37 Corey EugeneKay: Ah, I've seen those myself.
16:45 __gotcha_ joined #salt
16:46 Corey EugeneKay: I don't think it's init script based, I think it's the way we're calling the yum API
16:46 EugeneKay Could be. Like I said, I haven't even smoke-tested this with cmd.run
16:46 Corey EugeneKay: https://bugzilla.redhat.com/show_bug.cgi?id=861559
16:46 Corey I think we're screwed until upstream fixes it
16:47 EugeneKay Go figure.
16:55 aranhoide joined #salt
16:55 alekibango joined #salt
17:25 pentabular joined #salt
17:42 s0undt3ch iMil: you're iMilnb in github?
17:43 kmwhite joined #salt
17:45 __gotcha_ joined #salt
17:45 __gotcha_ joined #salt
17:50 Heartsbane joined #salt
17:50 Heartsbane joined #salt
17:55 the_drow joined #salt
17:55 the_drow How do I run a python script on all installed pythons?
17:56 the_drow They are all placed under /usr/bin/
17:57 the_drow I'd like to install pip using the get-pip script for all the python versions I installed
18:02 akoumjian joined #salt
18:03 akoumjian joined #salt
18:10 jeddi the_drow: do you want to install the python-pip package (or similar) do you mean?
18:11 jeddi or you want to install the latest pip *using* pip?
18:11 the_drow jeddi: no. I want to use get-pip.py
18:11 jeddi right.
18:11 the_drow https://raw.github.com/pypa/pip/master/contrib/get-pip.py
18:11 jeddi as a one-off?  you can just do cmd.run for the minions ... so:  salt '*' cmd.run '/usr/bin/get-pip.py'
18:11 jeddi that should do it - but test it, of course.  :)
18:12 the_drow jeddi: nope. I'd like to have it in my manifests
18:12 jeddi aha.  :)
18:12 the_drow jeddi: So, how can it be done?
18:13 jeddi the_drow: good question - i've only used the cmd. stuff for one-off's on the command line.  in a staet file you'll just need to have a qualifier like 'unless: ' where you check for the presence of some file (so that you don't keep running this script on every higihstate).
18:14 jeddi the_drow:  see http://docs.saltstack.com/ref/states/all/salt.states.cmd.html
18:14 jeddi so you might do cmd.run:  \\ - unless: 'ls /usr/bin/pip' \\
18:14 jeddi or wherever the pip binary gets installed to - is it /usr/local/bin ?
18:14 the_drow jeddi: or write a module?
18:15 jeddi writing a module seems a bit excessive to my mind, given it's only 3 lines in a state file to enforce this.
18:15 jeddi but .. i'm biased.  i haven't written any modules.  maybe they're easier than i think.
18:15 the_drow It's on /usr/local/bin/pip
18:16 Newt[cz] joined #salt
18:16 davisamo joined #salt
18:17 jeddi cool - well, does that statefile approach look like it would satisfy?
18:17 felixhummel joined #salt
18:18 the_drow jeddi: yes but not entirely. I want it to detect which pythons are installed and run it on all of them. Keeping the main as pip and the rest with pip2.6 etc.
18:19 jeddi the_drow: oh.  okay.  you could push another bash script out there that sensibly returned 0 / 1 based on your various conditions.  have that file.managed as a prerequisite for your pip cmd ... but it's getting to 6 lines of state file now, and an accompanying 5 line bash script. :)
18:20 jeddi the_drow: that's the way i'd do it because it's easiest and fastest for me.  clearly there are other ways, some may well be better.
18:20 jeddi the_drow: i take the view that 'get it working today - when you're bored, make it nicer'
18:24 Newt[cz]1 joined #salt
18:24 __gotcha_ joined #salt
18:46 Katafalkas joined #salt
18:50 Newt[cz] joined #salt
18:51 __gotcha_ joined #salt
18:52 iMil s0undt3ch: I am :)
18:52 iMil s0undt3ch: I've just replied to the issue
18:52 s0undt3ch iMil: Hello there! Thanks!
18:53 iMil you're very welcome, good to know I'm now alone on pkgin+salt :)
18:53 __gotcha_ joined #salt
18:53 __gotcha_ joined #salt
18:53 s0undt3ch iMil: well, I need to test the salt bootstrap script on SmartOS :) http://bootstrap.bb.ufsoft.org/waterfall
18:54 iMil wow, really cool
18:55 s0undt3ch iMil: yeah, SmartOS bootstraps but I need to run the test suite in it, hence my recent fixes
18:55 iMil s0undt3ch: so as I said on GH, pkgin 0.4 is really old and has nasty bugs in it
18:55 krissaxton joined #salt
18:55 s0undt3ch iMil: well, there doesn't seem to be a new one yet
18:56 iMil hmm, really surprising, Filip from Joyent maintains a pretty up-to-date package tree
18:56 s0undt3ch iMil: pkgin ug pkgin -> nothing to do.
18:56 iMil you're probably pointing to an old repo
18:57 iMil wait a sec
18:57 waverider joined #salt
18:57 s0undt3ch iMil: well, I spinned the VM up and didn't change anything else
18:57 iMil can you cat /opt/local/etc/pkgin/repositories.conf ?
18:58 iMil I'm using that repo: http://pkgsrc.joyent.com/packages/SmartOS/2013Q1/i386/All/
18:58 iMil which is the latest available AFAIK
18:58 s0undt3ch http://pkgsrc.joyent.com/sdc6/2011Q3/gcc46/All
18:58 iMil (2013Q2 enters the freeze period tomorrow)
18:58 iMil yeah
18:58 iMil like I said, old as fsck
18:59 s0undt3ch that's what joyend is using on their vms
18:59 iMil they actually ship their VM with that?
18:59 s0undt3ch yep
18:59 iMil :/
18:59 iMil well that sucks if you ask me...
18:59 s0undt3ch iMil: on the website
18:59 s0undt3ch ie
18:59 s0undt3ch on the cloud
18:59 iMil really old software in there
18:59 s0undt3ch don't know if downloading an ISO also uses that
19:00 iMil you definitely should open an issue on their GH
19:00 s0undt3ch darn
19:00 s0undt3ch now I notice
19:00 iMil I know them, they behave quickly
19:00 s0undt3ch 2011!?!?!?
19:00 iMil yeah..
19:00 s0undt3ch ouch
19:02 iMil actually when they gave me a virtual machine in order to check pkgin's behaviour on SmartOS is was setup woth the correct repo IIRC
19:03 iMil it*
19:03 iMil with*
19:03 * iMil looks at his fingers
19:03 s0undt3ch iMil: well, dunno what happened then
19:04 s0undt3ch https://github.com/joyent/smartos-live/issues/230
19:04 iMil nice
19:04 s0undt3ch iMil: Thanks! Need to get some dinner now :)
19:04 iMil enjoy :)
19:11 felskrone joined #salt
19:11 Newt[cz]1 joined #salt
19:14 pentabular joined #salt
19:15 cleeming[foxx] joined #salt
19:45 akoumjian joined #salt
19:45 akoumjian joined #salt
19:46 akoumjian joined #salt
19:46 akoumjian joined #salt
19:50 Ryan_Lane joined #salt
19:50 danieljn joined #salt
19:54 akoumjian joined #salt
19:55 linjan joined #salt
19:56 JordanRinke Morning
19:58 bhosmer joined #salt
20:00 mgw joined #salt
20:01 __gotcha joined #salt
20:14 yidhra joined #salt
20:17 krissaxton joined #salt
20:20 akoumjian joined #salt
20:26 akoumjian joined #salt
20:27 akoumjian joined #salt
20:27 akoumjian joined #salt
20:29 th3rt joined #salt
20:30 clintberry joined #salt
20:31 Gifflen joined #salt
20:31 krissaxton joined #salt
20:32 Gifflen joined #salt
20:36 davisamo joined #salt
20:40 th3rt left #salt
20:40 clintberry joined #salt
20:47 davisamo_ joined #salt
20:47 akoumjian joined #salt
20:48 bhosmer joined #salt
20:50 s0undt3ch iMil: still around?
20:53 Newt[cz] joined #salt
21:02 davisamo left #salt
21:07 bhosmer_ joined #salt
21:10 Politoed joined #salt
21:13 Gifflen_ joined #salt
21:15 cleeming[foxx] if anyone has a moment: https://github.com/saltstack/salt/issues/5567
21:26 agj joined #salt
21:27 krissaxton joined #salt
21:29 insatsu joined #salt
21:34 logix812 joined #salt
21:37 davisamo_ So,  when I try to put multiple unique providers in /etc/cloud or /etc/cloud.providers,  salt-cloud is always using the last provider definition.  My provider definitions include specific regions,  so this doesn't work well if salt-cloud is using only the last provider.  I am using the latest git of salt-cloud.
21:38 davisamo_ I've tracked the "problem" back to config.py and how it returns is_provider_configured().
21:39 davisamo_ The pydoc for is_provider_configured() says     ''""Check and return the first matching and fully configured cloud provider configuration. """
21:40 davisamo_ So is_provider_configured() always returns only what it thinks is the first match for the provider (in my case the provider is openstack).
21:44 davisamo_ The openstack.vm_ variable whose value is returned by openstack.get_configured_provider() is used  to authenticate against the endpoint and region returned by config.is_provider_configured()
21:45 davisamo_ So,  am I doing something wrong,  or is this a bug that I need to file?
21:46 jrgifford_ joined #salt
21:50 Asphyxiated|Dk joined #salt
21:50 Asphyxiated|Dk joined #salt
21:50 alekibango davisamo_: if it doesnt work up to your expectations, file the bug report, please
21:55 bhosmer joined #salt
22:02 kermit joined #salt
22:12 kmwhite joined #salt
22:16 clintberry joined #salt
22:30 krissaxton joined #salt
22:37 bfrog joined #salt
22:38 bfrog hi, looking at salt as a way of doing config management
22:39 bfrog wonder how it differs from something like puppet or chef?
22:39 bfrog found both of those kind of difficult to work with
22:40 bfrog can I just create shell like scripts and template some config files?
22:40 bfrog or am I going to have to learn a bunch of asanine terminology overload
22:51 mackstic1 Depennds on your definition of terminology overload
22:51 krissaxton joined #salt
22:51 bfrog recipes, chef, cookbook, blah blah
22:52 bfrog and also trying to do too much, like half the chef recipes have source installs for things
22:55 mackstic1 You probably won't like Salt much then from a terminology perspective, as it has about the same amount
22:57 mgw joined #salt
23:00 bfrog what about usage? is it pretty easy to do what I'm talking about
23:01 bfrog basically need to install some packages, run some stuff to setup things in a shell "postgres initdb ..." etc, modify/replace some config files, enable startup stuff etc
23:09 mackstic1 Yes, it'll do all that. You could even ignore 99% of the functioality and only use the file distribution and execution mechanisms if you really want to
23:12 faust joined #salt
23:19 Deece joined #salt
23:25 krissaxton joined #salt
23:26 cxz joined #salt
23:28 faldridge joined #salt
23:31 jeddi bfrog: i think salt attracts lots of chef and puppet refugees - you're in good company.  less terminology .. not sure.  as mackstic1 says there's a goodly amount.  but unlike chef and puppet, IMHO, you don't need to know it all at the outset.  there are no custom DSL's - for what you're talking about you can get away with YAML and nothing but YAML.
23:42 krissaxton joined #salt
23:43 adotbrown joined #salt
23:46 faust joined #salt
23:56 elbaschid joined #salt
23:58 __number5__ joined #salt

| Channels | #salt index | Today | | Search | Google Search | Plain-Text | summary