Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2014-09-22

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

All times shown according to UTC.

Time Nick Message
00:01 sectionme joined #salt
00:03 yomilk joined #salt
00:08 ravenac95 joined #salt
00:09 otter768 joined #salt
00:12 mpanetta joined #salt
00:18 Ryan_Lane joined #salt
00:30 mrlesmithjr joined #salt
00:31 mrlesmithjr joined #salt
00:34 ndrei joined #salt
00:46 mrlesmithjr joined #salt
00:59 elfixit joined #salt
01:00 mrlesmithjr joined #salt
01:02 sectionme joined #salt
01:02 malinoff joined #salt
01:13 bhosmer joined #salt
01:17 mpanetta joined #salt
01:19 patrek_ joined #salt
01:19 bhosmer_ joined #salt
01:23 aparsons joined #salt
01:24 aparsons_ joined #salt
01:30 goodwill joined #salt
01:32 XenophonF hey has anyone here built salt from source on windows?
01:42 mosen joined #salt
01:45 talwai joined #salt
01:45 talwai Is there a separate option in the master config if you're specifying multiple renderers? I'm using the line 'renderer: jinja | yaml | gpg' as specified here: http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.gpg.html . But getting the error [CRITICAL] The renderer jinja | yaml | gpg is unavailable, this error is often because the needed software is unavailable
01:46 bhosmer joined #salt
01:51 __number5__ talwai: do you have python-gnupg installed on that master?
01:55 talwai __number5__ , yes installed it. i'm still on v2014.1.5 , i'm beginning to think it's a version issue
01:58 rypeck joined #salt
02:03 sectionme joined #salt
02:21 Ajax2 joined #salt
02:29 dccc joined #salt
02:30 anotherZero joined #salt
02:38 ravibhure joined #salt
02:39 ravibhure left #salt
02:48 ramishra joined #salt
03:02 NotreDev joined #salt
03:04 sectionme joined #salt
03:08 bhosmer joined #salt
03:18 TyrfingMjolnir joined #salt
03:18 mgw joined #salt
03:23 vbabiy joined #salt
03:24 ramishra joined #salt
03:33 mrlesmithjr joined #salt
03:36 snuffeluffegus joined #salt
03:47 picker joined #salt
03:51 Ryan_Lane joined #salt
03:53 scoates joined #salt
03:57 active8 joined #salt
04:04 sectionme joined #salt
04:10 mrlesmithjr joined #salt
04:21 kingel joined #salt
04:25 NotreDev joined #salt
04:26 jeddi joined #salt
04:30 jalaziz joined #salt
04:35 ramishra joined #salt
04:37 jalaziz joined #salt
04:41 TyrfingMjolnir joined #salt
04:44 ravenac95 joined #salt
04:52 felskrone joined #salt
04:52 NotreDev joined #salt
04:53 jalaziz joined #salt
04:53 ramteid joined #salt
04:57 bhosmer joined #salt
05:03 jalaziz joined #salt
05:04 davet joined #salt
05:05 sectionme joined #salt
05:21 oyvjel joined #salt
05:25 davet joined #salt
05:26 kingel joined #salt
05:34 kingel joined #salt
05:50 sectionme joined #salt
05:58 catpigger joined #salt
06:08 sectionme joined #salt
06:09 colttt joined #salt
06:10 colttt joined #salt
06:19 Valdo joined #salt
06:23 madduck joined #salt
06:28 n8n joined #salt
06:28 mechanicalduck_ joined #salt
06:30 intellix joined #salt
06:31 lcavassa joined #salt
06:41 bhosmer joined #salt
06:46 bhosmer joined #salt
06:46 sectionme joined #salt
06:47 lionel joined #salt
06:58 slav0nic joined #salt
06:58 Sweetshark joined #salt
07:00 oyvjel joined #salt
07:12 jhauser joined #salt
07:13 anitak joined #salt
07:17 alanpearce_ joined #salt
07:23 briner hi, Is there a way to do a import my_lib in a state.sls written in #!py ?
07:25 mosen whats it for briner? sounds like it might be better to write an exection module
07:25 mosen sorry state module
07:25 aparsons joined #salt
07:27 aparsons_ joined #salt
07:36 martoss joined #salt
07:38 babilen briner: See http://docs.saltstack.com/en/latest/ref/modules/ and http://docs.saltstack.com/en/latest/ref/states/writing.html respectively
07:38 darkelda joined #salt
07:41 babilen briner: Also salt['yourexecution_mode.function'](arg1, arg2, ...) -- http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.jinja.html#calling-salt-functions
07:41 babilen s/mode/module
07:41 martoss1 joined #salt
07:46 che-arne joined #salt
08:00 totte joined #salt
08:08 tinuva joined #salt
08:11 floWenoL joined #salt
08:12 iamtew good morning
08:18 PI-Lloyd joined #salt
08:28 viq joined #salt
08:29 stephanbuys joined #salt
08:29 iclebyte joined #salt
08:30 akafred joined #salt
08:32 davidone joined #salt
08:32 kiorky joined #salt
08:33 aparsons joined #salt
08:35 bhosmer joined #salt
08:40 aparsons joined #salt
08:42 aparsons joined #salt
08:45 aparsons joined #salt
08:47 aparsons joined #salt
08:50 aparsons joined #salt
09:09 ramishra joined #salt
09:14 dccc joined #salt
09:19 halfss joined #salt
09:19 giantlock joined #salt
09:24 VSpike winrepo.genrepo outputs "failed to compile" errors but gives no clue at all what the problem is. Can anyone give me a clue? https://bpaste.net/show/f8b1639aef6b
09:25 oyvjel joined #salt
09:26 ninkotech__ joined #salt
09:29 VSpike I can import the init.sls file with python yaml, so it's not a low-level syntax issue
09:34 fredvd joined #salt
09:42 yomilk joined #salt
09:44 VSpike Well, that's one way to solve it. Extract a bit of http://fossies.org/dox/salt-2014.1.5/winrepo_8py_source.html to a local .py file and run it adding more instrumentation until I figure out the problem, which was adding "-" where there didn't oughta be none
09:49 wr3nch joined #salt
09:52 briner sorry for the late response. In fact, I use often #!py, and I'd like to push some code which are common for some states, in a central file.
09:52 ndrei joined #salt
09:59 anitak joined #salt
10:00 CeBe joined #salt
10:00 wnkz joined #salt
10:18 bhosmer joined #salt
10:19 stephanbuys joined #salt
10:23 bhosmer joined #salt
10:24 bhosmer_ joined #salt
10:28 Outlander joined #salt
10:31 aparsons joined #salt
10:40 stephanbuys joined #salt
10:45 mpanetta joined #salt
10:47 peters-tx joined #salt
11:01 TyrfingMjolnir joined #salt
11:03 g3cko joined #salt
11:08 scottpgallagher joined #salt
11:08 anitak joined #salt
11:10 stephanbuys joined #salt
11:19 micko joined #salt
11:23 hobakill joined #salt
11:24 younqcass joined #salt
11:25 diegows joined #salt
11:26 Daviey joined #salt
11:29 viq joined #salt
11:30 mrlesmithjr joined #salt
11:44 mindlessdemon joined #salt
11:47 stephanbuys joined #salt
11:51 StDiluted joined #salt
11:57 StDiluted anyone awake in here? :)
12:00 spo0nman joined #salt
12:05 TyrfingMjolnir joined #salt
12:06 Outlander joined #salt
12:06 PI-Lloyd StDiluted: nope :p
12:07 StDiluted hehe..
12:07 StDiluted I think i figured out my issue :)
12:08 StDiluted is Helium the upcoming release?
12:09 PI-Lloyd yup
12:09 StDiluted ok. So the ng formulas need that release to function, it appers
12:09 StDiluted appears*
12:11 g3cko joined #salt
12:11 PI-Lloyd ng formula?
12:11 StDiluted https://github.com/saltstack-formulas/nginx-formula
12:11 StDiluted for example
12:11 StDiluted “Note: nginx.ng requires the merge parameter of salt.modules.pillar.get(), first available in the Helium release."
12:11 PI-Lloyd ahh nginx, sry ng could of been anything, had to be sure :)
12:11 StDiluted well, no, there are ng formulas available for many of the formulas
12:12 StDiluted just so happens that nginx starts with ng too
12:12 PI-Lloyd oO this is the first I'm hearing of these formulas
12:12 bhosmer joined #salt
12:12 * PI-Lloyd goes off to poke around
12:13 PI-Lloyd "first available in the Helium release."
12:13 StDiluted yeah
12:13 stephanbuys joined #salt
12:16 mrlesmithjr joined #salt
12:16 mindlessdemon joined #salt
12:17 mindlessdemon joined #salt
12:17 PI-Lloyd you have another example of it, to me it looks like a formatting/structure thing, the thing those states are looking for is the "merge" function which is coming in Helium.
12:18 mindlessdemon joined #salt
12:18 StDiluted lemme see if i do
12:18 mindlessdemon joined #salt
12:18 PI-Lloyd yeah it's just a structure thing... notice how the folders are layed out... nginx/ng/init.sls
12:19 StDiluted yeah
12:19 PI-Lloyd so what you are actually looking for is the feature in Helium, the "merge" function/parameter
12:19 mindlessdemon joined #salt
12:19 StDiluted yep, that’s what I figured
12:19 StDiluted it’s ok
12:19 StDiluted I can use the standard one for now
12:20 StDiluted I just wasn’t sure if Helium was the next release or not
12:20 StDiluted or if i was just missing something
12:20 PI-Lloyd Yeah Helium is the next release
12:20 mindlessdemon joined #salt
12:21 TyrfingMjolnir joined #salt
12:22 jaimed joined #salt
12:24 ndrei_ joined #salt
12:26 stephanbuys joined #salt
12:33 TheRealBill joined #salt
12:36 StDiluted lol, do you need help, Guest?
12:40 felskrone anyone familiar with debugging failed tests? specifically this one: integration.wheel.key.KeyWheelModuleTest.test_list_all
12:41 spo0nman does anyone have experience in creating a highly available salt master?
12:41 johtso joined #salt
12:41 babilen multi master setup?
12:41 felskrone depends: there are two approaches: http://docs.saltstack.com/en/latest/topics/tutorials/multimaster_pki.html http://docs.saltstack.com/en/latest/topics/tutorials/multimaster.html
12:42 vejdmn joined #salt
12:43 spo0nman I've read both approaches, but i want to create a setup that's Multi-DC
12:43 aparsons joined #salt
12:45 bhosmer joined #salt
12:45 spo0nman i've read that you can create a master of master using salt-syndic, however syndic can only listen to one master, hence it makes the MasterofMasters a single point of failure.
12:46 felskrone multimaster syndic does not yet exist afaik
12:46 felskrone nor does syndic-multimaster-pki
12:46 spo0nman idea is to create 2 master of masters (one in each DC)  and then create DC specific masters to which the minions connect
12:47 aparsons joined #salt
12:47 spo0nman felskrone: yes. from the documentation, I gather that.
12:47 spo0nman however, Can i create a salt-syndic loop? Master1<->Master2 ?
12:48 lcavassa joined #salt
12:48 spo0nman and if I connect a minion to two masters, each of which runs syndic to a single MasterofMasters, do the commands execute twice on the minions?
12:49 jslatts joined #salt
12:50 StDiluted can someone look at this salt-cloud error and tell me what might be going on? https://gist.github.com/dginther/bff2278c01e77f1332e3
12:51 StDiluted machine gets created, but seems like the salt-cloud is erroring out, and no high state is executed
12:52 felskrone spo0nman: honestly, that sounds like a really complicated approach, not very pythonic considering the kiss principle :-)
12:54 teebes joined #salt
12:56 spo0nman felskrone: i think so too. So how do I do it? create a highly available salt master that can manage multiple DCs without minions connecting cross-dc
12:57 spo0nman that's the real question, i guess.
12:59 ericof joined #salt
13:00 miqui joined #salt
13:01 felskrone spo0nman: well, how would you prevent cross-dc minions with a single-master (being a multimaster setup)?
13:01 bhosmer joined #salt
13:03 spo0nman my idea was to create a master-of-masters (M1) and DC specific masters (DC1-M1) (DC2-M1) and make the DC specific masters run a syndic to subscribe to (M1)
13:04 spo0nman I can have multiple DC specific masters (so it's redundant)
13:04 babilen Which would mean that one of the DC specific master is accessing the MoM in another DC, wouldn't it?
13:04 spo0nman the only problem I have now is to make the Master of Masters redundant
13:04 spo0nman babilen: yes.
13:05 spo0nman that can be tolerated, however tens of thousand machines connecting cross dc is not tolerable
13:05 babilen I haven't seen a setup like this. Write to the mailing list as that might have been solved already (or not).
13:06 spo0nman babilen: i've written, https://groups.google.com/forum/#!topic/salt-users/_66cz2Qjzz4
13:06 spo0nman perhaps I can clarify a bit further.
13:06 aparsons joined #salt
13:10 babilen I understand what you are trying to do, but I don't think it is supported (yet)
13:10 felskrone spo0nman: i dont see a way to acomplish this with a syndic as as long as the syndic does not support multiple masters
13:11 aparsons_ joined #salt
13:11 spo0nman babilen: felskrone, thanks. I just wanted to confirm that my understanding was correct and I wasn't overlooking something obvious
13:12 spo0nman i also think it's not possible to remove the single point of failure in the Master of Masters, unless syndic supports multiple masters
13:12 spo0nman however, I was working on some work arounds, let me make some graphs and I'll take the groups opinion on what
13:13 spo0nman is the better work around
13:14 felskrone well, what you might be able todo is MoM(2+ servers with Multimaster-PKI-setup) <->  DC1(M1+M2 with MoM as maasters) <-> DC2(M1+M" with Mom as masters)
13:15 ghartz joined #salt
13:15 felskrone hm wait, then the syndic is missing again :-)
13:15 felskrone sorry,i gotta go
13:16 aparsons joined #salt
13:17 aquinas joined #salt
13:17 aquinas_ joined #salt
13:21 vbabiy joined #salt
13:21 racooper joined #salt
13:22 crisdev joined #salt
13:22 mpanetta joined #salt
13:30 StDiluted any assistance with a salt-cloud issue: https://gist.github.com/dginther/1d50d95f8d4f778c3c60
13:31 aparsons joined #salt
13:33 nitti joined #salt
13:33 PI-Lloyd StDiluted: Try deploying without the FQDN, and also try putting quotation around the hostname in the salt-cloud command
13:33 StDiluted ok
13:34 halfss_ joined #salt
13:34 StDiluted is it getting confused with the xxx.xxx.xxx?
13:34 StDiluted interesting
13:34 PI-Lloyd possibly, I ran into a similar problem with salt-cloud and GCE
13:35 StDiluted ok
13:36 cpowell joined #salt
13:37 micah_chatt joined #salt
13:37 dccc joined #salt
13:37 SheetiS joined #salt
13:39 babilen joined #salt
13:39 VSpike I guess this is still a live issue? https://github.com/saltstack/salt/issues/13513#issuecomment-46373007
13:39 VSpike I have powershell code already to add the appropriate keys so I can re-use that. It means you lose the state awareness though
13:40 dude051 joined #salt
13:40 sysadmin75 joined #salt
13:40 PI-Lloyd also StDiluted: what version of salt master are you running?
13:42 cxz joined #salt
13:43 cxz hey guys, just a note that the docs for releases don't extend properly with versioning: the 'next topic' for 2014.1.11 shows 2014.1.2
13:43 cxz http://docs.saltstack.com/en/latest/topics/releases/2014.1.11.html
13:43 cxz additionally 2014.1.9 points next to 2014.7.0
13:44 StDiluted I’m running Hydrogen, PI-Lloyd
13:44 cxz which is RC, right_
13:44 mechanicalduck joined #salt
13:44 cxz ?
13:45 PI-Lloyd StDiluted: hmm, then the only other cause would be an EC2 API change.
13:45 PI-Lloyd in which case you'd be better of filing a bug report
13:45 StDiluted trying with no FQDN
13:45 StDiluted seems like a bug report might be in order anyway
13:46 StDiluted you should be able to specify whatever you want for the node name
13:46 PI-Lloyd indeed you should
13:46 PI-Lloyd hence my thinking that the API has changed
13:47 StDiluted yeah it still did the same with no FQDN and the name of the node in single quotes
13:47 PI-Lloyd hmm, I can't find anything on the Amazon site regarding any API changes recently
13:47 mechanicalduck joined #salt
13:49 perfectsine joined #salt
13:50 micah_chatt_ joined #salt
13:50 stephanbuys joined #salt
13:52 Deevolution joined #salt
13:52 claytron joined #salt
13:58 halfss joined #salt
14:02 ramishra joined #salt
14:04 micah_chatt joined #salt
14:10 n1ck3 joined #salt
14:10 n1ck3 Good day poeple!
14:12 ajprog_laptop joined #salt
14:13 n1ck3 So I wrote yesterday about pkg.installed failing on OS X. I just wanted to write again and say that the issue seems to be isolated to OS X 10.10 Yosemite. I tried the same thing on a 10.9.5 Mavericks box and it worked without issue. I have posted an issue on Github (https://github.com/saltstack/salt/issues/16012). Just wanted to write that here, should someone else run across this issue.
14:15 dstufft joined #salt
14:16 dstufft Question, I'm looking to use the state module.wait to call the service.restart execution module, however the service.restart module has a "name" kwarg that I need to pass to it, but I don't think I can because the state has a "name" kwarg already
14:16 dstufft I'm not sure how to handle this :/
14:17 manfred dstufft:  m_name
14:17 manfred will be the name: variable for the module
14:18 dstufft ah, so module.wait: - name: <thing I want to restart> - m_name: service.restart ?
14:18 manfred no
14:18 manfred name: service.restart
14:18 manfred m_name: nginx
14:18 manfred m_name is the name variable you pass to the module
14:18 ckao joined #salt
14:18 manfred salt-call service.restart name=<m_name>
14:18 dstufft ah
14:18 dstufft oops
14:19 dstufft ok thanks
14:19 manfred np
14:19 rypeck joined #salt
14:24 rallytime joined #salt
14:26 aparsons joined #salt
14:30 slav0nic joined #salt
14:31 dude051 joined #salt
14:36 dude051 joined #salt
14:39 bhosmer_ joined #salt
14:39 dude051 joined #salt
14:42 dude051 joined #salt
14:46 mpanetta joined #salt
14:47 kermit joined #salt
14:47 the_drow joined #salt
14:48 the_drow Are you guys aware of salt modules/states that work with coreos?
14:48 the_drow Except etcd
14:48 the_drow salt-cloud doesn't support it yet
14:49 ajolo joined #salt
14:51 jeffspeff how can i get a minion to see that the masters dns ip has changed without restarting the minion service?
14:51 manfred jeffspeff: i don't believe you can, because the minion has an active connection opened to the master
14:52 PI-Lloyd jeffspeff: I believe the only way is to restart the minion
14:52 csa_ /me
14:54 jeffspeff here's what i'm doing... I have windows users with the minion installed on their laptops. when they're in the office the master's hostname resolves to the internal IP, but I also want it to work when they're outside of the office as well. I have the public dns pointing to the public IP and it works, but only after I manually restart the minion service.
14:56 kingel joined #salt
14:56 PI-Lloyd jeffspeff: in that case the best option is to have the minion always point to the public address, regardless of network connection
14:56 jeffspeff Ok, thanks PI-Lloyd
14:57 kusams joined #salt
15:00 VSpike jeffspeff: you could add a local dns override to avoid the NAT loopback too
15:00 dude051 joined #salt
15:00 VSpike assuming that's a thing
15:01 bhosmer joined #salt
15:01 viq VSpike: DNS override? As in it resolves to a different IP when minion is inside? He's already doing that
15:01 PI-Lloyd well ya, the internal DNS would have to be updated to point to the public address
15:02 VSpike oh sorry, totally missed that
15:02 VSpike yeah, I didn't think that through .. soz :()
15:02 PI-Lloyd lol
15:02 conan_the_destro joined #salt
15:03 blackhelmet Anyone have a second to step through the salt mine with me?
15:03 jeffspeff so, any suggestions on how to access the external ip of my office from within the office?
15:04 oz_akan joined #salt
15:05 PI-Lloyd jeffspeff, DNS entry should do it
15:06 PI-Lloyd depending on how your routing is setup ofc
15:06 cpowell joined #salt
15:07 jeffspeff PI-Lloyd, we're using a cisco asa 5510, i can change the dns to reflect the external IP, but I can't access the outside interface from the inside interface
15:07 viq jeffspeff: set up "multi"-master ?
15:09 PI-Lloyd jeffspeff: you could try dropping the local DNS and have it do the DNS lookup from global DNS...
15:10 viq PI-Lloyd: that doesn't help if he can't reach external address from the inside
15:10 sectionme joined #salt
15:10 PI-Lloyd I was under the impression the asa wasn't setup to do interface bridging/routing
15:12 Damon joined #salt
15:13 ggoZ joined #salt
15:17 anotherZero joined #salt
15:17 jalbretsen joined #salt
15:17 VSpike How can I tell if the win_gitrepos is working? For example, when I do salt-run winrepo.genrepo should it show the remote items as well as local ones?
15:18 eunuchsocket joined #salt
15:22 dude051 joined #salt
15:23 VSpike doh... helps if you read the docs properly
15:23 halfss joined #salt
15:27 colttt joined #salt
15:37 UtahDave joined #salt
15:37 wendall911 joined #salt
15:38 kingel joined #salt
15:38 smcquay joined #salt
15:39 smcquay joined #salt
15:39 tligda joined #salt
15:39 ekristen joined #salt
15:44 micah_chatt joined #salt
15:44 StDiluted joined #salt
15:45 jergerber joined #salt
15:45 felskrone joined #salt
15:51 xDamox joined #salt
15:53 perfectsine_ joined #salt
15:56 felskrone UtahDave: got a hint for me on this one https://github.com/saltstack/salt/pull/16019
15:56 eunuchsocket joined #salt
15:56 felskrone dont quite know what to do with it :-)
15:56 * UtahDave looking
15:56 felskrone damn you're fast
15:57 aparsons joined #salt
15:58 aparsons_ joined #salt
15:58 StDiluted UtahDave: when you’re done with that if you can take a gander at this: https://gist.github.com/dginther/1d50d95f8d4f778c3c60 I’d greatly appreciate it. Trying to use salt-cloud with command line: salt-cloud -p base_ec2_private ‘thing.atpay.com’
15:59 UtahDave felskrone: rallytime is looking at your issue right now.
16:00 felskrone awesome, just need to know what to do, once i get that, i'll fix it :-)
16:00 scbunn joined #salt
16:00 UtahDave felskrone: Yeah, rallytime will get you fixed up. She's awesome at the testing infrastructure.
16:01 felskrone she will probably grin a bit about the simplicity of the problem :-)
16:01 analogbyte joined #salt
16:02 UtahDave StDiluted: can you include your profile?
16:02 StDiluted utahdave: sure. one sec
16:03 troyready joined #salt
16:03 StDiluted i put it in the comment
16:03 thayne joined #salt
16:03 kingel joined #salt
16:03 StDiluted obviously it has indents in the original
16:04 StDiluted the VM gets created, i just get this error
16:04 kusams joined #salt
16:05 UtahDave StDiluted: if you add     ``` on the line before and after your code it will maintain the indents correctly
16:05 UtahDave what version of Salt are you on?
16:05 StDiluted Hydrogen
16:05 VSpike UtahDave: do you know if this will likely be fixed in the next release https://github.com/saltstack/salt/issues/13513 ?
16:06 VSpike Wondering how much effort to put into gold-plating my work-arounds
16:06 diegows joined #salt
16:07 UtahDave StDiluted: try changing your tag setting to this:  tag: '{"Environment": "production", "Role": "app_server"}'
16:08 UtahDave VSpike: Hm. I'll try to get that fixed this week.
16:08 kimberlad joined #salt
16:10 StDiluted UtahDave: giving that a try. Thanks.
16:10 VSpike UtahDave: oh awesome, thanks.. In that case, I'll keep the workarounds quick and dirty
16:11 UtahDave VSpike: I can't make any promises I'll get it done, though.  :)
16:11 wnkz joined #salt
16:12 StDiluted UtahDAve: same error.
16:12 VSpike Of course :)
16:12 ajprog_laptop joined #salt
16:12 XenophonF hey UtahDave, when you get a sec, I need a hand with building Salt on Windows
16:12 UtahDave StDiluted: can you add the output of    salt-master --versions-report
16:13 UtahDave XenophonF: what's going on?
16:13 XenophonF bdist_esky is throwing NameError
16:13 UtahDave XenophonF: can you pastebin the command and error?
16:14 XenophonF UtahDave: https://bpaste.net/show/4420bc00b9d3
16:14 anotherZero joined #salt
16:14 XenophonF command is simply python setup.py bdist_esky
16:14 StDiluted added comment, UtahDave
16:14 sectionme joined #salt
16:14 UtahDave XenophonF: Oh, I think that error was fixed in 2014.1.11
16:14 XenophonF so this is against the v2014.1.10 tag
16:15 XenophonF this is after running python setup.py install per http://docs.saltstack.com/en/latest/topics/installation/windows.html
16:15 UtahDave XenophonF: Yeah, that's a known problem. If you build 2014.1.11 that problem is fixed.
16:15 XenophonF OK thanks
16:16 UtahDave I can't remember which commit fixed it, but if you can find it you could probably cherry pick it.
16:16 UtahDave But I'd recommend just building 2014.1.11
16:16 XenophonF nah easier if i just go with the new ver
16:16 XenophonF thanks for the clue
16:16 UtahDave you're welcome
16:17 XenophonF is uninstalling salt a matter of deleting Python27\lib\site-packages\salt*?
16:18 XenophonF looks like there's stuff in Python27\scripts, too
16:18 UtahDave XenophonF: Yeah, that's it.
16:18 XenophonF good thanks again
16:19 XenophonF hm, maybe i should setup a virtualenv on windows...
16:19 elfixit joined #salt
16:20 ndrei joined #salt
16:20 ndrei_ joined #salt
16:20 UtahDave StDiluted: I'm not sure what's going on there.  Could you open an issue on that?
16:21 StDiluted UtahDave: sure. glad to.
16:21 UtahDave thanks!
16:23 debian112 joined #salt
16:23 KyleG joined #salt
16:23 KyleG joined #salt
16:23 StDiluted opened.
16:25 anitak joined #salt
16:29 aparsons joined #salt
16:30 kingel joined #salt
16:31 drawks hey all
16:31 drawks can someone explain to me really quick what the version numbering strategy is for salt?
16:32 analogbyte joined #salt
16:32 drawks I was about to add a warn_until call in a deprecated function, but I realize I don't clearly understand which version will be 2 releases from now
16:32 jcockhren drawks: year.month.bug_fix
16:32 jcockhren year.month is when that version of salt reached rc
16:33 babilen drawks: http://docs.saltstack.com/en/latest/topics/releases/version_numbers.html
16:33 drawks thanks
16:33 babilen (http://www.saltstack.com/salt-blog/2013/10/27/salt-version-numbers for some background)
16:34 drawks so when deprecating a function the warn_until should use a codename instead of a version number?
16:35 drawks or is just putting a date "sufficiently far" into the future the thing to do?
16:36 higgs001 joined #salt
16:36 drawks ah, cool I found a helpful comment in salt/version.py
16:36 drawks NAMES = {
16:36 drawks # Let's keep at least 3 version names uncommented counting from the
16:36 drawks # latest release so we can map deprecation warnings to versions.
16:38 VSpike where is the local cache for the git win repo on the saltmaster?
16:38 mo__ joined #salt
16:39 VSpike I've changed my config to point at a fork I made, and it says https://bpaste.net/show/91da0837c6e5 but it doesn't tell me where to do it
16:40 mo__ anyone on here using the latest release candidate? i'm interested in using salt to manage my vsphere installation
16:42 kingel joined #salt
16:43 mo__ okay; more information: i'm interested in using salt-cloud to deploy new vms from a base image snapshot; however, the vsphere provider doesn't support cloning from a snapshot of an image. adding this would be about 5-10 lines; is anyone else on here interested in this?
16:43 babilen drawks: http://docs.saltstack.com/en/latest/topics/development/deprecations.html
16:44 mo__ if i made the patch and submitted a pull request for the release candidate, would it be ignored until release candidate becomes full release?
16:45 manfred no?
16:45 manfred if you are adding a feature, it might make it into the next RC, but it depends
16:46 babilen drawks: So you'd use "salt.utils.warn_until("$CODENAME", "$REASONORMESSAGE")"
16:47 manfred the release candidate is pretty much locked into the features it is going to have for this release, mostly bug fixes are just being added now, your changes probably wouldn't make it in until lithium
16:47 mo__ manfred, okay. i guess i'll just make the change locally until it gets pulled. thanks
16:47 babilen mo__: Please develop against the "develop" branch, open a PR, and ask to have it backported into 2014.7 if you think that is sensible.
16:48 babilen (but I wouldn't necessarily expect new features to be added)
16:48 babilen Ah, manfred elaborated on that already.
16:48 mo__ babilen, thanks for the pointer about the develop branch
16:48 manfred new features should be made against develop, bug fixes against the current tagged branch it fixes the bug in.
16:48 babilen mo__: http://docs.saltstack.com/en/latest/topics/development/contributing.html has more information on htat.
16:49 babilen *that even
16:49 kiwnix joined #salt
16:51 micah_chatt joined #salt
16:52 clarkperkins joined #salt
16:52 NotreDev joined #salt
16:56 mo__ thanks, guys!
16:56 conan_the_destro joined #salt
16:59 apergos so... I"m having a pretty weird problem.  We deploy salt via puppet (no wisecracks form the peanut gallery please), ubuntu trusty with 0.17.5+ds; the package is installed, then puppet comes along and puts the proper config in place, and starts up puppet properly
16:59 apergos ter salt.
16:59 apergos this always fails; we get a whine that the master has rejected the public key
16:59 apergos I've done ths by hand (install the package, shoot the salt that starts up, copy in our good config)
16:59 apergos made ure, that there is no key on themaster hanging around and none on the minio
16:59 apergos n
17:00 apergos and when I invoke puppet it does  status  and then start which in the end calls salt-minion
17:01 apergos and we get this issue
17:01 apergos I check and the master has indeed got a different key thn the minion has written out
17:01 apergos we're not in a vm or anything, no cloud, no funny filesystem
17:02 apergos if I delete that key from the master then it's all fine on the next puppet run
17:02 apergos what I want to know is, what culd be causing salt to generate a key, and then generate a new one in the space of, well, a few to no seconds I guess
17:02 rallytime @felskrone did you see my comment on your PR?
17:04 SheetiS apergos: did you write your own puppet manifest, or are you using one publicly available?
17:04 jslatts joined #salt
17:04 apergos we have our own manifest because we manage rather a lot
17:04 apergos manifests, I should say
17:04 apergos the salt piece is pretty simple though
17:06 apergos I've checked via lastcomm that salt-minion only gets executed once, so it's not that
17:07 hobakill i did something stupid. i issues a cmd.run 'service salt-minion restart' to a bunch of my linux boxes and now they're not responding to a test.ping
17:08 Ryan_Lane joined #salt
17:08 SheetiS is it just install the package and ensure=>running on the service?  I'd want to rule out anything puppet was doing first (Just like I'd want to rule out things inside a salt state when troubleshooting it).
17:08 mwillman_ joined #salt
17:08 apergos well we are well past installing the package
17:08 apergos I have the package installd nd our good conf file in place
17:08 apergos nothing running
17:09 apergos I run puppet, the only thing it needs to do is start the service
17:09 apergos which it duly does...
17:09 apergos and startig with no key for the minion on the master, and no keys on the minion either, we get the reject
17:10 clarkperkins I'm having an odd issue when setting salt as a dependancy in a setup.py file.  We have 'salt==2014.1.10' listed as one of our requirements, but here's what happens when we pip install our package:
17:10 clarkperkins Downloading/unpacking salt==2014.1.10   Downloading salt-2014.1.10.tar.gz (2.8MB): 2.8MB downloaded   Running setup.py egg_info for package salt     2014.1.0     package init file 'salt/templates/__init__.py' not found (or not a regular file)   Requested salt==2014.1.10, but installing version 2014.1.0
17:10 clarkperkins Has anyone ever seen that pop up before?
17:13 SheetiS apergos: and puppet is not managing anything directly in /etc/salt/pki for the master or the minion, correct?  (when the puppet agent runs, the only change is just the service { "salt-master": ensure => "running", } triggers or whatever?)
17:13 apergos that is correct
17:13 apergos we don't preseed or any of that stuff, no shipping around of any funny keys
17:14 sectionme joined #salt
17:15 Gareth morning morning
17:16 rihannon joined #salt
17:16 anitak joined #salt
17:17 SheetiS Are the both basically starting at the same time?  Maybe there is something funny that goes on when salt-master starts the first time if the minion connects immediately as it is starting?  I'm totally guessing at this point.
17:18 Ryan_Lane apergos: the master is failing to start?
17:18 hobakilllll joined #salt
17:18 Ryan_Lane is this after an upgrade?
17:18 Ryan_Lane or still on 0.17?
17:19 rihannon I have a jinja template in which I want to do something like:    This_year = {{ datetime|strftime('%Y') }}
17:19 rihannon The documentation says this is possible, but I can't get it to work (datetime doesn't exist).
17:19 rihannon How can I get the current time using jinja?
17:19 apergos master is fine, Ryan_Lane
17:19 apergos this is the minion, on trusty, with 0.17.5
17:19 Ryan_Lane ah
17:19 Ryan_Lane is the version of zmq the same?
17:20 Ryan_Lane 0mq, that is
17:21 mwillman_ joined #salt
17:23 apergos it was some reasonable one
17:23 apergos ok I'm going to assume the data has to be coming from somewher else
17:23 apergos this is ridiculous
17:24 StDiluted is it generally considered best practice to remove anything in top.sls that you have ‘included’ into another state that is in the top.sls, or does it really not make a difference?
17:24 mwillman_ Hi I am in the process of building out a SALT installation for our Comapny, Can anyone point me in the direction of some good examples of how to begin organizing states and services.
17:25 StDiluted rihannon, did you try something like: {{ car.date_of_manufacture.strftime('%Y-%m-%d') }}
17:25 SheetiS rihannon: did you try {{ "now"|strftime('%Y') }}
17:25 Ryan_Lane apergos: what's the issue again?
17:25 UtahDave StDiluted: It doesn't really matter, but it can make your top.sls simpler and shorter to remove states that are included elsewhere.
17:26 Ryan_Lane I don't have the backscroll
17:26 StDiluted UtahDave: ok. as long as it doesnt matter i like the readability of having it in the top.sls, I think
17:27 StDiluted thanks
17:27 apergos master has some other key than the minion is sending... or the minion sends two keys.... or who knows.  anyways, the master rejects the key and I have to manually remove it out of th emaster pki store, then on the next run everythingn is fine
17:27 UtahDave clarkperkins: yeah, that's a known problem on 2014.1.10. Can you bump that to 2014.1.11?
17:29 felskrone rallytime: i did. if i understood it correctly, i would just put all changes (salt as well as salt-testing) in a single PR and the test would all pass, right?
17:29 UtahDave apergos: The minion will generate its own key upon startup if a key doesn't already exist
17:29 apergos yes.
17:29 rallytime felskrone yep!
17:29 apergos that's what it shoul do.  once.  and send it to the master,
17:29 martoss joined #salt
17:29 rallytime you could probably just pull the testing changes into the original PR, if you wanted to felskrone
17:30 Ryan_Lane apergos: on the minion remove all of /etc/salt/pki
17:30 aparsons joined #salt
17:30 mwillman_ Hi I am in the process of building out a SALT installation for our Comapny, Can anyone point me in the direction of some good examples of how to begin organizing states and services?
17:31 rallytime felskrone you mind need to make a few other changes. I just just checked my test run and there's still a few things that might need to be adjusted.
17:31 UtahDave mwillman_: just a sec
17:31 Ryan_Lane hell, remove all of /etc/salt, purge the minion package and reinstall it
17:31 clarkperkins @UtahDave thanks!  yes, we can use 2014.1.11 instead
17:31 Ryan_Lane apergos: then rerun puppet
17:31 rihannon SheetiS:  SaltRenderError: Jinja error: Unable to parse now. Consider installing timelib
17:31 rihannon I guess I'll look into installing timelib…
17:32 rallytime felskrone the changes would be very similar to the change you made in integration.wheel.key where the expected out put needs to be adjusted
17:32 felskrone rallytime: yes, there are more failures i will have to figure out. but having to do all the changes in a single was my missing piece in the puzzle :-)
17:32 rallytime does that make sense?
17:32 rallytime yep, mine too at first :)
17:32 apergos Ryan_Lane: this is a repeatable failure; it happens on every trusty install
17:32 Ryan_Lane apergos: ahhh, ok
17:32 felskrone rallytime: awesome, thanks for your help :-)
17:32 Ryan_Lane I bet the PKI directory changed
17:32 rallytime anytime felskrone!
17:32 apergos if I install the package by hand and
17:33 apergos put our good minion config in place
17:33 kusams joined #salt
17:33 apergos and make sure the pki dir is empty on the minion
17:33 clarkperkins @UtahDave I just tried 2014.1.11 and I get the same error
17:33 apergos nd that the minions-pre dir has not got any key on the master, running puppet will give me this headache
17:33 Ryan_Lane apergos: we actually set most of the options in the config file
17:33 Ryan_Lane this is my error
17:33 apergos after the config file sis already in place, see...
17:33 Ryan_Lane we should have a standard /etc/salt/minion.conf
17:34 apergos yeah well it's been chopped down to just a few lines now
17:34 rihannon StDiluted: That requires that I already have a time coded into the car variable.  I don't want to do something like {% set year = 2014 %} for obvious reasons.
17:34 Ryan_Lane then only set the explicit options we want in /etc/salt/minion.d/wiki.conf
17:34 apergos and the pki dir isn't one of them
17:34 Ryan_Lane ah, so it's the default?
17:34 apergos but even if it were...
17:35 SheetiS rihannon: probably need timelib on the machines that would get now that way (pip install timelib would likely fix or manage with a state)
17:36 rallytime felskrone FWIW the other failures you're going to hit are in the integration.shell.key file.
17:37 felskrone great, thx :.)
17:38 rallytime NP!
17:38 xDamox joined #salt
17:40 UtahDave mwillman_: Here's a gist with the tree from  my /srv/salt on my testing Salt Master.  https://gist.github.com/UtahDave/665c086547b44cb3ebe1
17:41 debian112 so I am having pillar problem here: I can pull the pillar data down via: salt-call pillar.get server1_gt
17:41 debian112 http://paste.debian.net/122423/
17:41 UtahDave mwillman_: I like to keep related sls files together in a directory.
17:41 debian112 but I am unable to populate a jinja template with it.
17:42 UtahDave debian112: can you pastebin the related part of your sls file?
17:42 debian112 sure thing
17:42 apergos Ryan_Lane: so I was saying, even if there was a pki line in the ocnf file, the conf file is already in place when I do this testing
17:43 hobakill joined #salt
17:43 apergos here's the state of things: no running salt minion. /etc/salt/pki/minion empty on minion. /etc/salt/pki/master/minions_pre empty on master (and no other version of minion key in the other dirs either)
17:44 apergos package installed, our goo config of 5 lines in place
17:44 apergos then I puppet gent test, the only thing tht happens is
17:44 oyvjel joined #salt
17:44 apergos Notice: /Stage[main]/Salt::Minion/Service[salt-minion]/ensure: ensure changed 'stopped' to 'running'
17:45 apergos and I go to check the log: the salt master has rejected this key... I check the master, it now has a "pre" key with the current timestmap tot he minute anyways, and a different m5d than the one (also current to the minute) the minion has
17:45 apergos 100% repeatable
17:46 Ryan_Lane apergos: I bet the minion is starting with a default configuration when the minion starts, then is being restarted and for some reason is changing its hash
17:46 Zachary_DuBois joined #salt
17:46 Ryan_Lane likely being reconfigured
17:47 debian112 @UtahDave: http://paste.debian.net/122425/
17:48 UtahDave debian112: address: {{ salt ['pillar.get']('interface_eth0:address') }}
17:48 UtahDave try that instead.
17:50 felskrone rallytime: just tried it locally and everything's passing. you rock! :-)
17:50 apergos well it's not being restarted from the command line, because lastcomm does't show more than one salt-minion being fired off
17:50 rallytime felskrone awesome!
17:50 felskrone UtahDave: that also goes out to you, as always been a great help :-)
17:50 debian112 @UtahDave: pillar data here: http://paste.debian.net/122430/
17:50 UtahDave rallytime++
17:50 aboe joined #salt
17:51 aparsons joined #salt
17:51 hobakill UtahDave: is there a  way to reliably restart all the salt-minion service on windows?
17:51 UtahDave hobakill: what version are you on?
17:51 * hobakill fails english
17:51 hobakill 2014.1.10
17:52 UtahDave debian112: then this:     address: {{ salt ['pillar.get']('server1_gt:interface_0:address') }}
17:53 UtahDave hobakill: if you upgrade to 2014.1.11 then you can do    salt 'windowsid' service.restart salt-minion
17:53 jslatts joined #salt
17:53 UtahDave hobakill: otherwise you'll have to do this with powershell:      salt 'windowsid' cmd.run 'Restart-Service salt-minion' shell=powershell
17:53 hobakill as soon as it hits epel i will. thanks UtahDave
17:54 jalaziz joined #salt
17:54 XenophonF UtahDave, where is the code that determines the minion ID?
17:54 UtahDave hobakill: You can just download the windows installer.
17:55 hobakill OH yeah i suppose...windows...derp. thanks.
17:55 perfectsine joined #salt
17:55 apergos Ryan_Lane: puppet claims to execute "/sbin/start salt-minion"... when I do this by hand, from the same state of empty pki dirs etc, it works, i.e. I see the master has cached the key and wants me to sign
17:55 ericof joined #salt
17:55 apergos but from a puppet run, fail every time
17:57 Ryan_Lane apergos: hm. I wonder if it's actually calling service start salt-minion
17:57 Ryan_Lane apergos: is there a script in /etc/init.d/salt-minion and in /etc/init/salt-minion?
17:58 bhosmer joined #salt
17:59 debian112 @UtahDave I have that: http://paste.debian.net/122425/
17:59 debian112 address: {{ salt ['pillar.get']('server1_gt:interface_0:address') }}
17:59 UtahDave debian112: no, in that paste you have    interface_eth0
18:00 UtahDave not    interface_0
18:00 mwillman_ thanks
18:02 StDiluted is anyone using salt and AWS autoscaling together?
18:02 debian112 @UtahDave: wow I am an idiot
18:02 debian112 can't believe I over looked that
18:03 debian112 @UtahDave thank you!
18:03 UtahDave :)  I was starting to doubt my eyes!   :)
18:03 debian112 work fine
18:03 yetAnotherZero joined #salt
18:03 aparsons joined #salt
18:05 ndrei joined #salt
18:05 ndrei_ joined #salt
18:05 chitown_ .win move 9
18:06 scbunn joined #salt
18:07 arnoldB UtahDave: could you guys add the google mailing list salt-users-de@googlegroups.com to the web page http://www.saltstack.com/community/?
18:09 UtahDave arnoldB: Yeah, for sure! thanks!
18:10 arnoldB going to post to salt-users@ and a g+ posting. sharing the g+ posting could be also nice
18:10 apergos Ryan_Lane: it's calling start:
18:10 apergos 1) that's in lastcomm output, 2) --debug shows
18:10 jmccree joined #salt
18:11 apergos Debug: Executing '/sbin/start salt-minion'
18:11 apergos answers may be  a little delayed because meeting.
18:13 anitak joined #salt
18:14 oz_akan joined #salt
18:15 sectionme joined #salt
18:16 apergos the idea that it is somehow is looking elsewhere initially for its config is intriguing (but then wouldn't it try to talk to 'salt', which doesn't exist... and fail to generate any keys at all?)
18:18 scbunn joined #salt
18:18 ndrei joined #salt
18:18 ndrei_ joined #salt
18:20 arnoldB UtahDave: https://plus.google.com/u/0/113495771554638768621/posts/WuLBgC5kyQp
18:21 arnoldB UtahDave: https://groups.google.com/forum/?hl=de#!topic/salt-users/4eR6oE2-tfo
18:22 babilen .. beigetreten.
18:23 ajolo joined #salt
18:24 ndrei joined #salt
18:24 ndrei_ joined #salt
18:30 bhosmer_ joined #salt
18:32 drawks anyone with a little more time on this project wanna give my PR a look-see for sanity?
18:32 drawks https://github.com/saltstack/salt/pull/16034
18:32 drawks something just didn't sit right with having a module to partition and format disks that didn't have even one test :)
18:32 babilen Would anybody tracking develop be so kind to try to reproduce #15968 ? I think that #7079 which was fixed before was re-introduced by mistake in c8c6112.
18:33 babilen (essentially somebody seems to have forgotten to rebase against develop before submitting the PR)
18:39 fannet joined #salt
18:40 crane hey ho, i'm trying to transform a static config file in a dynamic configuration with jinja. but i dont know how to pass the variables into my template file when using two or more minions. can someone help me out?
18:41 iggy crane: try paste'ing your files somewhere
18:42 fannet http://docs.saltstack.com/en/latest/ref/states/all/salt.states.gem.html  - I'm literally using this exact example with passing the gem name and salt keeps saying "Name addressable in slsgem is not a dictionary" .... anyone have any ideas?
18:42 ajolo joined #salt
18:43 crane iggy: but it is all very raw... https://github.com/craneworks/tor-salt
18:43 manfred fannet: can you paste your sls file into pastebin?
18:43 NotreDev joined #salt
18:43 jeffspeff how can i send a custom command to a minion? something like "salt '*' do.command'service network restart'"
18:43 crane iggy: i hope (and think) i need just a small example. for example how to define SocksPort for two different hosts
18:43 manfred cmd.run 'service network restart'
18:44 manfred or service.restart network
18:44 jeffspeff manfred, thank you very much
18:44 jaimed joined #salt
18:44 jeffspeff I'm very new to salt (less than a week) and I'm loving the project and the community.
18:46 bhosmer joined #salt
18:46 iggy crane: there are multiple ways you can do things like that... you can put variables in the pillar and target the different pillars at different hosts, you can have use if's in the template file and key off the grain nodename
18:46 fannet #manfred: http://pastebin.com/FT3hK6gM
18:46 iggy it just kind of depends on the rest of your setup, etc.
18:47 crane iggy: i gues pillars are the more secure way?
18:47 manfred fannet: gem.installed shouldn't be a list
18:47 iggy well, your pillars are in a public github repo, so...
18:47 crane iggy: it's a dummy setup ^^
18:47 iggy but generally speaking, yeah, pillars are for "secure" data
18:47 fannet manfred: Am I reading this wrong: http://docs.saltstack.com/en/latest/ref/states/all/salt.states.gem.html ?
18:48 manfred fannet:  http://ix.io/erc
18:48 crane iggy: could you give me a small example how to deal with this idea?
18:48 iggy crane: check just about everything in github.com/saltstack-formulas for various ways to use pillar data
18:49 kermit joined #salt
18:49 fannet Thanks manfred!  Trying now
18:49 crane iggy: ah nice, i will dig into this :)
18:50 perfectsine joined #salt
18:51 arnoldB crane: this guy also published some formulas: https://github.com/bechtoldt?tab=repositories
18:52 arnoldB though there are some hardcore formulas :D
18:54 kusams joined #salt
18:55 crane looks like the template makros are very useful? ^^
18:56 arnoldB crane: that always depends on the use case
18:57 crane arnoldB: i never used jinja before so this will be intresting
18:58 StDiluted Anyone using AWS autoscaling and salt-cloud in concert?
19:01 babilen crane: If you want to hit somebody better use Python or mako ;)
19:02 crane babilen: well... in fact that django is also using jinja... i think i have to learn it now
19:02 \ask joined #salt
19:02 jslatts joined #salt
19:02 nitti_ joined #salt
19:02 iggy it's not exactly the same, but close
19:02 babilen crane: http://jinja.pocoo.org/docs/dev/templates/ is what you want to read
19:03 kingel joined #salt
19:03 arnoldB when I need to build more complex/ deployment formulas I'd use the pyobjects (python) renderer
19:03 arnoldB doing db schema updates, etc.
19:03 crane babilen: thx, i will read it while i'm going home. enough work for today. :)
19:04 babilen arnoldB: I make liberal use of formulas which means that my pillars are typically more complicated than my states :D
19:05 BrendanGilmore joined #salt
19:06 Guest69574 joined #salt
19:06 iggy ^
19:09 arnoldB babilen: more complicated than https://github.com/bechtoldt/iscdhcp-formula/blob/master/pillar.example.sls ?
19:10 CeBe joined #salt
19:11 babilen arnoldB: Oh .. yes.
19:12 fannet sometimes when I spawn a DigitalOcean node I get Error: There was a profile error: Executing the command '/tmp/.saltcloud-146013c5-9593-41c4-8106-cf1d44aafe56/deploy.sh -c /tmp/.saltcloud-146013c5-9593-41c4-8106-cf1d44aafe56' failed and the creation fails. Most times it works fine. Any reason why this occurs?
19:12 manfred fannet:  it is a known issue in salt-cloud with transfering the file over, we haven't nailed it down 100%
19:12 manfred i have been unable to reliably replicate it, so it is difficult to trouble shoot
19:12 spookah joined #salt
19:13 manfred i have never been able to get it to happen with -D in script_args, and —keep-tmp and -l debug on salt-cloud
19:13 repl1cant joined #salt
19:14 bhosmer joined #salt
19:14 anitak joined #salt
19:16 nitti joined #salt
19:17 babilen arnoldB: https://www.refheap.com/c09fd36c32eef3b15cab34b30 would be an example.
19:21 bmcorser joined #salt
19:31 pancho_ joined #salt
19:31 pancho_ howdy
19:31 pancho_ i have a question regarding states.sls
19:31 bhosmer_ joined #salt
19:31 pancho_ I want to set dependencies between them
19:31 pancho_ as in don't run this state unless this other state has finished
19:31 pancho_ does anyone know how could I do that
19:31 babilen pancho_: http://docs.saltstack.com/en/latest/ref/states/requisites.html
19:31 KyleG joined #salt
19:31 KyleG joined #salt
19:32 iggy or orchestrate
19:32 arnoldB babilen: http 404
19:33 babilen arnoldB: Yes, I removed it.
19:33 pancho_ this provides dependencies per statements on a states.sls
19:33 arnoldB babilen: k
19:33 aparsons joined #salt
19:33 pancho_ I want to say don't run X state.sls unless Y has been ran and completed
19:34 pancho_ The only place I see to set a linear order is on the top.sls
19:34 bezeee joined #salt
19:37 ajolo joined #salt
19:37 joeyparsons joined #salt
19:37 babilen arnoldB: https://www.refheap.com/1456a04a9d53af6b208f7dbcd
19:38 babilen pancho_: You are looking for require/require_in
19:38 catpigger joined #salt
19:38 babilen pancho_: Ah, yes. Dependencies are per-state.
19:39 babilen pancho_: Could you elaborate why per-state dependencies are not suitable?
19:40 nitti joined #salt
19:41 arnoldB babilen: interesting approach. that gives me some ideas..
19:41 pancho_ maybe I am thinking about to linearly. I was think as finish my "package.sls" than my "application.sls" and build a state one on top of the other
19:41 babilen arnoldB: I am quite influenced by Clojure/List and always think "Want to write a TLD" :)
19:42 babilen pancho_: Why would that be necessary?
19:42 arnoldB pancho_: do you try orchestrating or do you want states to be executed in a specific order?
19:42 pancho_ orchestrating
19:43 babilen What are you *really* trying to achieve?
19:43 pancho_ reading the orchestrating doc, I think this is what I am looking for
19:44 arnoldB pancho_: look for Salt overstate, reactor, etc. but as babilen tried to ask: what is the exact process you try to realize?
19:45 babilen Yes, there are many approaches, but what pancho has said so far does not even necessitates enforcing a specific order of states let alone SLS files in total.
19:45 arnoldB ack
19:47 pancho_ first, I am segregating my states. I had a huge state file that did it all from lvm, packages to installing and deploying my application
19:48 babilen go on
19:49 pancho_ so I have some states that compile couple libraries and install specific software, that rely on say my package,sls and pip.sls
19:51 pancho_ I wanted to know if I could say, this state say "compile.sls needs package.sls and pip.sls, from what I am seeing each statement in my compile,sls should have the require: sls: -package.sls andn pip.sls
19:52 pancho_ and I apologize if I am approaching this the wrong way
19:52 babilen You really should define dependencies in a more granular way and, specifically, per-state. compile.sls does not require package.sls. Those are just meaningless files. *Some* states might require *some* packages in pacakges.sls, so make that explicit.
19:52 pancho_ ok thanks
19:53 babilen You can require: sls: package, but I wouldn't recommend to do that.
19:54 debian112 now that external_ip is gone, is there another way to get the external IP address?
19:55 jalaziz joined #salt
19:57 perfectsine joined #salt
19:58 iggy debian112: we have a custom gce grain that gets that (and pulls tags and other metadata into grains)
19:58 nitti joined #salt
20:00 jhulten joined #salt
20:02 * babilen steals iggy's crystal ball
20:03 kingel joined #salt
20:05 debian112 iggy you can to share?
20:06 debian112 or do I need wip out my python skills?
20:06 jcockhren debian112: pun?
20:06 jcockhren I kid!!
20:06 KennethWilke joined #salt
20:11 drawks ug... the lint pass of the jenkins build ruffles my feathers. it should still do the test run instead of dying right away. I'd like to see my failed units and the lint fail in the same run so i could try and fix them in a single commit
20:12 drawks also I'm not sure what linter this project uses, but damn is it strict
20:13 drawks I run flake8 and only ever submit code that is clean via flake8
20:13 UtahDave drawks: the test suite still runs even if the lint test fails.
20:13 drawks but somehow the jenkins lint run always find something
20:13 UtahDave Jenkins got backed up a little this morning, I believe
20:13 drawks ah
20:14 drawks it looked like it died after lint
20:14 UtahDave There is a  .pylintrc here: https://github.com/saltstack/salt/blob/develop/.pylintrc
20:15 quickdry21 joined #salt
20:15 UtahDave drawks: It's possible that there was some weirdness this morning when Jenkins was backed up, but every pull request gets a lint check and tests run on Ubuntu and Cent
20:16 drawks whats with the base64 blob in the pylintrc ?
20:16 drawks it decodes to something innocuous enough. why the opaque style?
20:17 intellix joined #salt
20:17 talwai joined #salt
20:17 bitmand joined #salt
20:18 UtahDave drawks: I have no idea. I'd have to ping s0undt3ch about that one.
20:19 nitti_ joined #salt
20:20 talwai How does one add a custom renderer to the salt master configuration? E.g. I have 'my_custom_renderer' under the _renderers/ directory in file_roots. The master knows about this renderer, which I can verify by running saltutil.sync_renderers. However when I add the following to my salt master config: renderer: jinja | yaml | my_custom_renderer , I get a renderer not available error. Any insight into this?
20:22 higgs001 joined #salt
20:24 UtahDave talwai: is that the correct syntax? Looks like you can just specify one renderer in that config option
20:26 talwai @UtahDave The composition syntax works for built-in Salt renderers, as detailed here: http://docs.saltstack.com/en/latest/ref/renderers/. E.g. If I use 'renderer: jinja | yaml' it works just fine. But I see no documentation on whether custom renderers can be added as part of the pipeline, such as 'jinja | yaml | custom_renderer'
20:27 UtahDave talwai: Hm. I think they should, but there might be a corner case where this isn't working correctly.  Would you mind opening an issue for this?
20:28 UtahDave talwai: also, what renderer are you adding?  I'm just curious
20:29 bitmand left #salt
20:30 eriko joined #salt
20:30 talwai @UtahDave, trying to add a basic decryption renderer, inspired by http://docs.saltstack.com/en/latest/ref/renderers/all/salt.renderers.gpg.html which isn't available in the Salt version I'm bound to.
20:30 UtahDave ah, ok.  cool.
20:30 smcquay joined #salt
20:30 talwai And yes will definitely open an issue, if I find this to be a bug on current develop as well
20:31 UtahDave you could probably just copy that gpg renderer to your install and use it
20:32 jslatts joined #salt
20:33 thayne joined #salt
20:33 talwai @UtahDave: Yep, I was going to try that out next
20:34 UtahDave cool
20:35 mgw joined #salt
20:37 iggy debian112: I can probably share, give me a few minutes to finish up something at work
20:38 drawks hehe nice, jenkins returning 502 now
20:43 UtahDave drawks: yeah, they're attending to it right now
20:47 moderation joined #salt
20:50 MISaxx joined #salt
20:54 forrest joined #salt
20:56 goodwill joined #salt
20:57 MISaxx left #salt
20:59 rawtaz is there anything like a roadmap or such where one can estimate progress on solt-ssh becoming up to date and fully compatible with all of salt features?
20:59 manfred rawtaz:  they removed the line marking salt-ssh as beta in the release candidate
20:59 manfred other than that… the goal was helium… no idea if they met it
21:00 manfred https://github.com/saltstack/salt/pull/15944
21:01 n8n joined #salt
21:02 rawtaz that os good news
21:02 rawtaz is*
21:05 kballou joined #salt
21:05 bhosmer joined #salt
21:05 rawtaz is fileserver operations the only thing not yet supported by salt-ssh?
21:06 manfred no idea, i rarely touch that part of salt :/
21:06 rawtaz :)
21:06 iggy debian112: http://pastie.org/9585859 GCE metadata grains
21:07 kusams joined #salt
21:07 yomilk joined #salt
21:07 UtahDave iggy: that's pretty awesome.  Would you be willing to add that to the grains in salt-contrib?
21:09 iggy yeah, I'm guessing that's just a matter of fork->send pull request?
21:09 iggy i.e. no kind of license forms or anything crazy
21:09 mechanicalduck_ joined #salt
21:09 UtahDave iggy: yep!   the repo is at    https://github.com/saltstack/salt-contrib
21:10 UtahDave That would be really awesome. Just Apache 2.0 license. You keep the copyright
21:11 iggy there's some other metadata you can grab too, that's just the stuff we use so far
21:14 micah_chatt joined #salt
21:15 rodo joined #salt
21:16 iggy gah! now I've got 3 things I need to finish today
21:16 iggy and only one of them is actually work
21:17 rawtaz reading about salt-ssh the other day, i found some text in the documentation that mentioned a few things that didnt work with salt-ssh. but i cant find that now. it was not http://docs.saltstack.com/en/latest/topics/ssh/index.html , it was another page (more elaborate on what things in the state file wouldnt work). any idea what page that might have been? im trying to establish what isnt supported with salt-ssh
21:18 yomilk joined #salt
21:24 avn joined #salt
21:24 sysadmin75 left #salt
21:26 n8n_ joined #salt
21:28 rawtaz bug report: in http://docs.saltstack.com/en/latest/topics/releases/releasecandidate.html the first paragraph's GitHub link isnt right
21:29 kusams_ joined #salt
21:29 eunuchsocket1 joined #salt
21:30 mo__ joined #salt
21:30 mo__ anyone out there use salt-cloud to deploy a windows vm?
21:30 mo__ i'm having trouble figuring out how to write the deploy script
21:30 mo__ any pointers?
21:31 rawtaz can you be more specific? what is it you dont understand about writing that deploy script?
21:31 bmonty joined #salt
21:31 hasues_ joined #salt
21:31 mo__ where is the installer going to be put?
21:32 rawtaz i cant help but lets see if anyone else can
21:32 mo__ do i have to install smbclient on the template?
21:32 hasues joined #salt
21:32 mo__ how do i make the windows template listen on 445?
21:32 mo__ okay
21:33 capricorn_1 joined #salt
21:33 __mel joined #salt
21:34 scooby2 joined #salt
21:35 __mel Hi. i have a sls usinf file.managed to configure my default grub config. now i have different server hardware and like to use different config option depending on the hardware. how would i do that?
21:36 __mel cut the config file into little pieces and build it up unpon some if the else statements?
21:43 avn joined #salt
21:43 iggy I can do a pillar.get foo and it comes back with the correct data, but in my formula (which uses the map.jinja merging with pillar data pattern), the pillar value is not making it into the template
21:43 beach joined #salt
21:43 iggy thought?
21:43 iggy thoughts?
21:43 beach does anyone know why gce.py was dropped out of https://github.com/saltstack/salt-cloud/tree/develop/salt_cloud/saltcloud/clouds
21:44 beach where can I get it again?
21:44 iggy good luck
21:44 manfred beach: we don't use that anymore
21:44 beach uhhh... why not. There is no longer support for gce?
21:45 manfred no, salt-cloud got merged into salt proper
21:45 manfred https://github.com/saltstack/salt/blob/develop/salt/cloud/clouds/gce.py
21:45 beach ahh... whoa... how did i miss that
21:45 manfred no idea :P happened in january, before the 2014.1 release
21:46 beach nah
21:46 racooper joined #salt
21:46 beach I've been using it for sometime... it seemed to break on my last update last week.... maybe since I tried to up date salt-cloud
21:46 beach hell, I don't even know the syntax to run cloud command from salt.
21:47 manfred still salt-cloud
21:47 iggy it's the same
21:47 manfred https://github.com/saltstack/salt/blob/develop/scripts/salt-cloud
21:48 kermit joined #salt
21:48 beach okay, sorry I think I'm confused. so we still use salt-cloud, but gce was dropped?
21:48 manfred https://github.com/saltstack/salt/blob/develop/salt/scripts.py#L282
21:49 iggy the package formerly known as salt-cloud has been merged into upstream salt
21:49 iggy it has gce support
21:49 manfred no
21:49 manfred it all still is supported
21:50 iggy (although in my opinion, it's pretty buggy)
21:50 manfred just don't install salt-cloud from pip, use the salt-cloud that comes in regular salt)
21:50 manfred iggy:  pfft
21:50 iggy just sayin...
21:50 beach should I remove salt-cloud
21:50 manfred did you install it using pip?
21:50 beach sorry for all the questions, I feel so out of the loop and I try to keep up to date with salt
21:51 beach no, I thought I installed with apt (debian)
21:51 beach I may have, I installed and started using salt-cloud quite some time ago
21:51 iggy should be pretty easy to tell
21:51 vejdmn joined #salt
21:51 iggy which salt-cloud
21:52 iggy if it's in /usr/local, it's a pip install
21:52 beach yeah, pip freeze says I have salt-cloud installed (0.8.11)
21:53 beach it's in /usr/local
21:53 mo__ any windows salt-cloud users?
21:53 iggy so yeah, pip uninstall that thing
21:54 mo__ especially vsphere?
21:55 mechanicalduck joined #salt
21:56 VSpike UtahDave: thanks for accepting the msysgit pull request. are there any guidelines to what would be accepted in the saltstack win repo github? I have a few other things to install
22:01 beach oh gawd yes it's working
22:01 aparsons_ joined #salt
22:03 beach okay, trick question then.... any idea on how to add in scopes (equivilant to IAM roles in EC2) in a gce based profile?
22:03 jalaziz joined #salt
22:03 johtso joined #salt
22:04 ericof joined #salt
22:05 jalaziz joined #salt
22:06 murrdoc joined #salt
22:07 hasues When discussing bugs in the install_salt script, would the salt-dev channel be more appropriate?
22:08 murrdoc you could do it here
22:09 murrdoc then submit a pull request
22:10 hasues Okay, I'll look at setting up what's necessary for the pull.
22:11 apergos oh a salt-dev channel... htm
22:19 acabrera joined #salt
22:23 analogbyte joined #salt
22:24 mo__ VSpike, do you use salt-cloud with vsphere to launch vms and deploy salt?
22:26 UtahDave VSpike: As far as guidelines for the winrepo, we're pretty open to including anything that people would find useful. So please go ahead and send any PR's you'd like to see in there.
22:26 saltmoli joined #salt
22:27 vejdmn joined #salt
22:28 KennethWilke joined #salt
22:29 saltymoli joined #salt
22:30 kingel joined #salt
22:31 aparsons joined #salt
22:33 aparsons joined #salt
22:34 Outlander joined #salt
22:35 ekristen UtahDave: so I’m trying to replicate the git pillar issue and do some upgrade testing too — have you ever seen a problem were only the first ext_pillar git will clone out into /var/cache/salt/master/pillar_gitfs?
22:37 UtahDave I'm not sure.
22:37 UtahDave are you waiting the 60 seconds for the refresh?  What versions of Salt are you testing?
22:39 ekristen UtahDave: so i’m currentlying OLD code 2014.1.0rc3 because I wanted to dupe my environment, test it and then upgrade and test again
22:40 ekristen but I can’t get my pillar data to work properly, I’m wondering if this is the same bug I ran into when I tried upgrading to 2014.1.7 the first time, only the first pillar remote is getting pulled, and I’m willing to bet the systems I saw all my pillar data disappear on was not in that pillar’s environment
22:40 ekristen I don’t want to waste time troubleshooting an older version but I am also curious as to why I can’t dupe my existing environment :/
22:41 UtahDave Yeah. Let me check something
22:42 vejdmn joined #salt
22:42 ekristen k, btw, my logging is at debug on the master and I see zero errors
22:44 UtahDave ekristen: does it seem to be like this issue?  https://github.com/saltstack/salt/pull/9992
22:45 ekristen let me double check, but I bet you that is exactly the problem
22:50 hasues When creating a masterless minion, the docs state to change the file_client directive.  What file is this located?
22:51 debian112 Thanks IGGY
22:51 higgs001 joined #salt
22:51 UtahDave hasues: /etc/salt/minion
22:53 hasues UtahDave: Thanks!
22:53 UtahDave You're welcome!
22:53 iggy debian112: keep your eyes peeled on the salt-contrib repo... I guess I'll be updating that with the rest of the metadata to go in there
22:53 ekristen UtahDave: not sure that it is — the first one in the list under ext_pillar, always gets checked out properly and if the minion belongs to that environment, all is well, regardless if pillar_roots exists or not
22:54 ekristen UtahDave: I’ve tried adding pillar_roots and entries for my two test branches pointing to an empty directory on the filesystem, but still doesn’t seem to fix it, the first in the list works, any others after that do not
22:54 UtahDave OK. Hm. I believe it updates/polls every 60 seconds. Have you waited the long between checkking?
22:55 UtahDave aahh. OK, can you pastebin that part of your config?
22:55 ekristen sure thing, one sec
22:55 ekristen and yes I always wait until the pillar sync happens
22:57 yomilk joined #salt
22:57 ekristen UtahDave: http://pastebin.com/rWtvmT21
22:58 ekristen UtahDave: /srv/pillar/base exists, writable by root, but is empty (perhaps that is the problem, that it is empty)
22:59 ekristen the dev branch works perfectly, branch4 however does not, if I remove pillar_roots all together and swap the order of branch4 and dev, branch4 works perfectly the others do not
22:59 ekristen I’m wondering if something in /var/cache/salt/master/pillar_gitfs got screwed up and if I should just blow that away and restart the master and see what happens
23:03 saltymoli anyone here work with the vsphere cloud provider? i'm specifically wondering why the developer decided to use samba, and a deploy script, instead of just using the pysphere send_file/start_process
23:04 ekristen UtahDave: so only the pillar_gitfs/0 directory gets touched
23:04 ekristen so weird
23:04 ekristen I’m concerned about how my production environment is working properly at the moment
23:06 dstufft btw is there any sort of ETA on when 2014.7.0 will go final? ;)
23:06 micah_chatt joined #salt
23:08 ekristen UtahDave: any idea when that bug was fixed? which version specifically?
23:08 ajolo joined #salt
23:09 rihannon left #salt
23:09 snuffeluffegus joined #salt
23:09 aparsons joined #salt
23:09 micah_chatt_ joined #salt
23:10 sherbs_tee joined #salt
23:11 UtahDave saltymoli: probably because they were pretty new to pysphere.  :)  that sounds easier.
23:11 UtahDave ekristen: have you tried using 2014.7.0RC2?
23:11 n8n joined #salt
23:12 UtahDave dstufft: We're planning on cutting hopefully the last RC later this week to send it to our QA hordes.
23:13 mosen joined #salt
23:13 dstufft UtahDave: ok cool
23:14 dstufft UtahDave: some stuff I want to use there ;) but I don't want to install stuff via pip into the system level on the PSF machines
23:15 kusams joined #salt
23:15 sherbs_tee is it expected to get a highstate failure when relative-including an empty state file?
23:15 UtahDave dstufft: Yeah, I bet!  People are always really excited to upgrade for new features, but I must always caution everyone to obey their good sysadmin instincts and test thoroughly before upgrading in production!  :)
23:16 UtahDave sherbs_tee: Hm. I don't think so. Can you pastebin the error you're getting?
23:16 Outlander joined #salt
23:17 jslatts joined #salt
23:18 sherbs_tee UtahDave: http://pastebin.com/M8KDECR1#
23:18 sherbs_tee thanks
23:19 sherbs_tee UtahDave: this is actually a masterless minion
23:19 sherbs_tee master is localhost
23:19 sherbs_tee if highstate runs fine if I don't include .cookie
23:19 UtahDave sherbs_tee: can you try adding something innocuous to that file to see if it runs?
23:20 sherbs_tee I tried with "---" and also "foo: bar", and both had the same result
23:20 sherbs_tee I can try once more
23:21 sherbs_tee UtahDave: same error after adding "baz: quux" to cookie.sls
23:23 seanz joined #salt
23:23 seanz Greetings everyone. Is there a "great" way to change the IP address of a minion?
23:23 seanz Through salt, that is.
23:23 UtahDave sherbs_tee: wait, is that a pillar file or state file?
23:23 sherbs_tee pillar
23:23 sherbs_tee sry, should have said tha
23:23 sherbs_tee t
23:24 sherbs_tee my state is simply installing a package
23:24 sherbs_tee ...which works if I remove the .cookie reference in the pillar init.sls
23:24 chrisjones joined #salt
23:25 UtahDave I wonder if relative includes don't work with pillars.   /me looking
23:25 rglen joined #salt
23:26 sherbs_tee UtahDave: bah -- https://github.com/saltstack/salt/issues/8875
23:26 UtahDave sherbs_tee: can you try giving the full path from the file_roots?
23:26 sherbs_tee sure...processing.....
23:26 ingwaem joined #salt
23:27 UtahDave yeah, good find on that issue. Looks like relative imports don't work with pillar.
23:27 seanz If I change the IP address of a machine and restart the minion on that machine, will the minion send out the new IP info to the master?
23:29 UtahDave seanz: the minion connects to the master, so the master doesn't really care about the minion's IP
23:29 ingwaem seanz, you can get the ip from the grains. Salt’s communication doesn’t rely on ips but rather machine names
23:30 ajolo joined #salt
23:30 sherbs_tee UtahDave: bah...how should I reference from the file_roots? you mean with salt://
23:30 sherbs_tee ?
23:30 ingwaem yea what dave said :)
23:30 scbunn joined #salt
23:30 UtahDave sherbs_tee: no.  What's the full path to that file?
23:31 sherbs_tee it's /srv/salt/salt-states/role/admin/go/cookie.sls
23:31 seanz UtahDave, ingwaem: I have a server that doesn't have a DNS record, yet salt can still connect to it with the fqdn. That tells me salt must be doing something there with the IP.
23:31 sherbs_tee s/states/pillar/g
23:31 sherbs_tee file_roots is defined as /srv/salt/salt-pillar
23:32 ingwaem seanz: you only need to define the masters ip in the minions config file, /etc/salt/minion …. since the minion establishes a persistent connection to the master, the master doesn’t need the ip of the minion to pass through commands. This helps salt maintain it’s fast communication
23:32 seanz ingwaem: Oh...so it's actually the minion that is connected to the master.
23:32 ingwaem seanz: you would start hitting problems if the master’s ip was changing all the time, so hopefully you can accommodate that
23:33 seanz I see - I wasn't aware of that fact.
23:33 UtahDave sherbs_tee: do you mean pillar_roots?
23:33 sherbs_tee UtahDave: yes, pardon my constant slip up of that
23:33 seanz So when I change the IP on the machine with the minion, I don't even have to restart the minion.
23:33 UtahDave so include:\n  - role.admin.go.cookie
23:33 ingwaem seanz: it shouldn’t be necessary
23:33 seanz Good to know!
23:33 sherbs_tee cool
23:33 sherbs_tee thanks
23:33 sherbs_tee processing...
23:34 ingwaem seanz: you can always try this out by viewing the grains of a minion before and after changing it’s ip
23:34 seanz Ah, a good test.
23:34 seanz I'll work that into my procedure.
23:34 ingwaem salt '*' grains.items
23:34 sherbs_tee UtahDave: that did the trick
23:35 UtahDave ah, good
23:35 sherbs_tee UtahDave: awesome...thanks. So I guess relative imports don't yet work with pillar
23:35 UtahDave sherbs_tee: no, they apparently don't
23:35 sherbs_tee cool...ok...as long as it's known
23:35 seanz ingwaem: Thanks.
23:36 ingwaem seanz: yw
23:37 mechanicalduck_ joined #salt
23:38 ajprog_laptop joined #salt
23:42 bezeee joined #salt
23:46 aparsons_ joined #salt
23:48 bill_h joined #salt
23:55 sherbs_tee left #salt
23:56 mechanicalduck joined #salt
23:57 rallytime joined #salt

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