Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2015-03-08

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

All times shown according to UTC.

Time Nick Message
00:08 iggy thornycrackers: allows salt-cloud to login and bootstrap the instance
00:10 thornycrackers I thought that specifying a key name would use ssh keys that you have entered onto your provider. How does it use that key to login to the minion? In examples it specifies a private key usually so I'm just confused on how it uses that key
00:14 iggy I'd have to go back and look closer at our setup, it's been a while since I looked at the configs
00:15 thornycrackers It's all good I was just curious if someone knew off the top of their head
00:16 Andre-B_ joined #salt
00:24 yomilk joined #salt
00:34 pahko joined #salt
00:43 ajw0100 joined #salt
00:48 yomilk joined #salt
00:50 amcorreia joined #salt
01:01 smcquay_ joined #salt
01:06 oblio joined #salt
01:07 IOMonster joined #salt
01:07 IOMonster joined #salt
01:11 dthorman joined #salt
01:12 I3olle joined #salt
01:13 thayne joined #salt
01:16 Phibs joined #salt
01:17 mafro joined #salt
01:29 cberndt joined #salt
01:31 thayne joined #salt
01:33 aphoriser joined #salt
01:47 cberndt joined #salt
01:48 Andre-B joined #salt
01:48 viq joined #salt
01:54 cotton joined #salt
02:07 aquinas joined #salt
02:07 dthorman joined #salt
02:11 nikogonzo https://github.com/saltstack/raet/blob/master/raet/lane/stacking.py#L87 <- this causes ioflo/base/aiding.py to try and setsockopt SO_SNDBUF to 6553500; thi causes socket.py to fail in openbsd with socket.error: [Errno 55] No buffer space available
02:11 nikogonzo does anyone know if this is an openbsd tunable? i have tried several sysctls and quickly glanced at man login.conf without success
02:13 cberndt joined #salt
02:14 cotton_ joined #salt
02:17 aphorise joined #salt
02:17 michelangelo joined #salt
02:21 yomilk joined #salt
02:27 murrdoc joined #salt
02:29 StDiluted joined #salt
02:29 cberndt joined #salt
02:32 catpiggest joined #salt
02:36 catpigger joined #salt
02:36 scoates joined #salt
02:55 helderco joined #salt
03:01 aphorise joined #salt
03:07 sjwoodr joined #salt
03:12 peters-tx2 Saturday night and I'm fiddling with Salt
03:12 peters-tx2 8/
03:17 mafro joined #salt
03:19 Aikar pff
03:19 Aikar ive been working 100+hr weeks for over a month :P
03:20 Aikar salt at 5am = yep
03:25 mdln joined #salt
03:26 mdln joined #salt
03:34 cberndt joined #salt
03:34 CeBe1 joined #salt
03:45 aphoriser joined #salt
03:49 active8 joined #salt
03:51 fragamus joined #salt
04:06 malinoff joined #salt
04:09 StDiluted joined #salt
04:14 aphorise joined #salt
04:15 StDiluted joined #salt
04:23 StDiluted joined #salt
04:24 peters-tx2 whiteinge, FYI I found a better "10.14. WHAT IS THE BEST WAY TO RESTART A SALT DAEMON USING SALT" FAQ solution, as I commented here https://github.com/saltstack/salt/pull/18186
04:24 micko1 joined #salt
04:24 germs_ joined #salt
04:28 dthorman joined #salt
04:43 ajw0100 joined #salt
04:52 germs_1 joined #salt
05:00 dmpeters63 joined #salt
05:02 aphorise joined #salt
05:06 dmpeters63 left #salt
05:08 markizano joined #salt
05:10 linjan joined #salt
05:14 fllr joined #salt
05:15 favadi joined #salt
05:23 dooshtuRabbit joined #salt
05:24 Ryan_Lane joined #salt
05:26 dooshtuRabbit joined #salt
05:27 thayne joined #salt
05:27 dooshtuRabbit joined #salt
05:33 aphorise joined #salt
05:43 randomuser joined #salt
05:45 otter768 joined #salt
06:13 markizano joined #salt
06:22 raygunsix joined #salt
06:32 peters-tx2 Dang RHEL7 systemd.
06:33 tkharju joined #salt
06:34 JoeJulian joined #salt
06:35 armguy joined #salt
06:39 evle joined #salt
06:48 peters-tx2 whiteinge, Hmm, I didn't realize the FAQ was in GIT; I'll attempt to log a pull request
06:49 linjan joined #salt
06:56 peters-tx2 whiteinge, Yay, pull request created
06:56 peters-tx2 https://github.com/saltstack/salt/pull/21410
06:57 MaliutaLap joined #salt
06:57 MaliutaLap left #salt
07:45 otter768 joined #salt
07:46 MohShami joined #salt
07:47 MohShami hey guys, I want to have a lot of pillar files in one folder, is it possible to include them all in top.sls like we do with states, or is that not possible?
07:47 jeddi joined #salt
07:48 iggy can you do that in state top.sls files?
07:49 MohShami iggy: yep, if I say for example, server.*
07:50 MohShami that includes all state files in /states/server
07:50 iggy that's the first I've heard of that
07:50 MohShami already have it running on 2 servers
07:52 linjan joined #salt
07:53 MohShami but for some reason that doesn't work for pillars
08:08 stooj joined #salt
08:37 nikogonzo https://github.com/saltstack/raet/blob/master/raet/lane/stacking.py#L87 <- this causes ioflo/base/aiding.py to try and setsockopt SO_SNDBUF to 6553500; thi causes socket.py to fail in openbsd with socket.error: [Errno 55] No buffer space available
08:37 nikogonzo does anyone know if this is an openbsd tunable? i have tried several sysctls and quickly glanced at man login.conf without success
08:39 chutzpah joined #salt
08:46 cbezawad joined #salt
08:52 dthorman joined #salt
08:57 bhosmer_ joined #salt
09:02 NV joined #salt
09:06 Fiber^ joined #salt
09:15 gfa joined #salt
09:19 ndrei joined #salt
09:28 badon_ joined #salt
09:28 ajw0100 joined #salt
09:31 scarcry joined #salt
09:36 scarcry joined #salt
09:46 otter768 joined #salt
10:11 fllr joined #salt
10:12 aquassaut joined #salt
10:16 Andre-B joined #salt
10:25 Singularo joined #salt
10:27 markm joined #salt
10:38 intellix joined #salt
10:49 packeteer joined #salt
11:01 netzmonster joined #salt
11:04 bluenemo joined #salt
11:06 felskrone joined #salt
11:06 I3olle joined #salt
11:09 packeteer joined #salt
11:19 mafro joined #salt
11:27 bhosmer joined #salt
11:28 stooj joined #salt
11:39 devweasel joined #salt
11:47 otter768 joined #salt
11:52 baffle joined #salt
11:54 smcquay_ joined #salt
11:54 _JZ_ joined #salt
11:57 __JZ__ joined #salt
12:00 fllr joined #salt
12:01 bash1245_ joined #salt
12:04 felskrone joined #salt
12:04 linjan joined #salt
12:07 dude051 joined #salt
12:09 Andre-B joined #salt
12:10 dude051 joined #salt
12:13 NV joined #salt
12:13 bash1245_ joined #salt
12:14 thehaven joined #salt
12:27 thayne joined #salt
12:45 JlRd joined #salt
13:08 pdayton joined #salt
13:09 bash1245_ joined #salt
13:20 catpig joined #salt
13:22 aphoriser joined #salt
13:36 favadi joined #salt
13:38 bhosmer joined #salt
13:48 otter768 joined #salt
13:49 fllr joined #salt
13:51 devweasel joined #salt
13:56 otter768 joined #salt
14:10 UmInAsHoE joined #salt
14:10 otter768 joined #salt
14:11 UmInAsHoE hi all, I'm struggling to find an answer to if you can populate a pillar value from a file content?
14:12 UmInAsHoE like: 'mykey: salt://mykey.value'
14:15 esharpmajor joined #salt
14:22 eliasp UmInAsHoE: https://github.com/saltstack/salt/issues/18406
14:22 thayne joined #salt
14:23 UmInAsHoE eliasp: ty!
14:33 triplecorn joined #salt
14:37 triplecorn joined #salt
14:39 eliasp does anyone of you use symlinks in GitFS repos for states? using symlinks seems to completely break state.py here… it just doesn't find any data targeted through symlinks… e.g. "include: \n - ntp" fails when ntp is a symlink in a repo pointing to "ntp-formula/ntp"
14:41 eliasp meh… it's not only state.py, but also the renderer… a Jinja include will fail just the same way ;-/
14:50 aphorise joined #salt
14:53 alexbst joined #salt
15:09 MTecknology eliasp: No! Don't be silly... A symlink is a Windows thing and not a git thing.
15:10 eliasp MTecknology: symlinks are supported just fine by git…
15:11 eliasp MTecknology: and I've used symlinks on Linux for a decade before I found out, windows (at least on NTFS) supports them too ;)
15:14 Kelsar Alternate Data Streams
15:14 Kelsar I wih we had them with linux
15:16 eliasp Kelsar: xattr?
15:16 Kelsar eliasp: not the same, you can only store a limited amount of data there
15:17 eliasp ah, so an ADS is actually like a full FS inode used for xattr with "unlimited" storage?
15:17 GabLeRoux joined #salt
15:17 Kelsar eliasp: simplyfied, yes
15:31 bluenemo_ joined #salt
15:32 teall joined #salt
15:35 fllr joined #salt
15:37 fllr_ joined #salt
15:37 brianfeister joined #salt
15:52 UmInAsHoE is it possible to have pillar state files have '.' in their filename and still include them? e.g. 'mycertificate.crt.sls'
16:08 clintberry joined #salt
16:18 ndrei joined #salt
16:25 bfoxwell joined #salt
16:25 johanek joined #salt
16:34 johanek joined #salt
16:36 fllr joined #salt
16:45 johanek joined #salt
16:59 StDiluted joined #salt
17:16 CheKoLyN joined #salt
17:18 fllr joined #salt
17:19 linjan joined #salt
17:21 smcquay_ joined #salt
17:30 StDiluted joined #salt
17:42 teall joined #salt
17:45 hasues joined #salt
17:50 hasues left #salt
17:55 bigpup3 joined #salt
18:00 chamunks joined #salt
18:02 mdln joined #salt
18:05 aquinas joined #salt
18:07 iggy nikogonzo: you should probably ask about ioflo/raet stuff somewhere else... it's not in wide use yet (but after saltconf, I hope that changes)
18:23 ldlework joined #salt
18:24 ldlework left #salt
18:30 fllr I'm creating a custom module, and I'm using the highstate outputter for that. How do I make salt change the color of that output to red in case I hit some error?
18:30 peters-tx2 Return a non-zero result code?
18:39 fllr peters-tx2: I'm already returning a dictionary in order to output using the highstate format
18:48 CeBe joined #salt
18:51 fllr Hey guys! I'm creating a custom module, and I'm using the highstate outputter for that. How do I make salt change the color of that output to red in case I hit some error?
19:00 iggy did you look to see how other modules do it?
19:01 fllr iggy: I'm trying to, but I haven't been able to figure it out...
19:02 fllr iggy: it's incredibly confusing what other modules do, and there's 0 documentation on how to do it
19:02 bash124512 fllr maybe you mean state ? I don't remember modules returning colors but I could be mistaken
19:03 fllr bash124512: No. I'm writing a module, and I'm trying to output using the highstate outputter...
19:09 bash124512 just an idea, but try returning false on an error. Never used the outputter b4.
19:10 fllr That would make no sense, since I'm trying to use the outputter to give me more information than just getting a false back :/
19:11 peters-tx2 fllr, maybe here?  https://github.com/saltstack/salt/blob/develop/salt/output/highstate.py#L142
19:12 fllr peters-tx2: Right... I've been playing with salt's code, and it seems execution stops at line 125, actually...
19:12 fllr I have no idea why...
19:12 fllr I think the problem is this line: sorted(data, key=lambda k: data[k].get('__run_num__', 0))
19:12 fllr But I can't figure out why that is being an issue...
19:14 sxar joined #salt
19:16 peters-tx2 fllr, Maybe your output isn't properly formatted?
19:16 fllr peters-tx2: Very likely. I'm trying to figure out the format right now. Like I said, this isn't documented anywhere...
19:17 peters-tx2 fllr, I'm pretty sure at the Conf they said everything is basically YAML, unless you override
19:19 fllr what?
19:19 mafro joined #salt
19:22 peters-tx2 fllr, if I run salt minion state.highstate -l all    Right before the highstate outputter spits out its info, I see "[TRACE   ] data ="  It has the actual data coming back from the minion.
19:23 peters-tx2 Including " 'result': True,"  Which determines the color the highstate outputter is going to use...
19:23 peters-tx2 I'm super newb but it seems logical
19:25 peters-tx2 fllr, I got LOTS more data if I go the the minion system and run   salt-call state.hightstate -l all
19:25 peters-tx2 fllr, That might be more useful to you... I see "[DEBUG   ] Results of YAML rendering:"
19:28 ckao joined #salt
19:32 ajw0100 joined #salt
19:37 bash1245_ joined #salt
19:41 ndrei joined #salt
19:51 amcorreia joined #salt
19:55 bash1245_ joined #salt
20:12 jhauser joined #salt
20:15 viq joined #salt
20:17 bigpup3 joined #salt
20:22 pf_moore joined #salt
20:23 Andre-B joined #salt
20:29 MatthewsFace joined #salt
20:30 germs_ joined #salt
20:38 cberndt joined #salt
20:45 gogolady joined #salt
20:49 brianfeister joined #salt
20:54 amcorreia_ joined #salt
20:59 garthk joined #salt
21:14 aphorise joined #salt
21:15 chitown_ joined #salt
21:19 nikogonzo iggy: thx, will try the github unless you can think of a better place
21:20 iggy nikogonzo: might be the best place
21:20 triplecorn joined #salt
21:21 chiui joined #salt
21:22 badon joined #salt
21:22 iggy fllr: I don't think you should be forcing a module to output using a specific outputter... you should use normal logging routines and allow salt to output using the outputter the user actually wants to use
21:23 peters-tx joined #salt
21:28 brianfeister joined #salt
21:33 triplecorn joined #salt
21:33 triplecorn_ joined #salt
21:39 andreyst joined #salt
21:45 fllr iggy: That's not the issue... I'm logging normally... it's just the results of my module i want to look nice... so it's easily digestible by the user...
21:45 fllr iggy: I'm following the documentation with this... It's just that this is really poorly documented...
21:51 peters-tx2 fllr, It looks simple to me.  Alter the Result code in the DICT that you are returning.  I've told you this before.  Here's another example https://github.com/saltstack/salt/blob/develop/salt/states/cmd.py#L675
21:52 peters-tx2 Salt CMD state will change the 'result' to False it the command is a failure.
21:52 peters-tx2 It will pack the failure output into 'comment'
21:52 peters-tx2 This will feed upwards to the state engine.
21:53 peters-tx2 It looks pretty basic to me
21:53 fllr peters-tx2: I left for my lunch break, sorry. Lol. I just now saw all the messages you sent :P
21:53 fllr peters-tx2: Though, I've tried setting 'result' to false
21:54 fllr It still shows green
21:54 fllr Pasting code should help...
21:54 helderco joined #salt
21:54 peters-tx2 fllr, It shows green AND it says "Result: False" ?
21:55 fllr peters-tx2: Yes.
21:56 fllr peters-tx2: https://gist.github.com/felipellrocha/ea407a31554db3badbbe
21:57 mosen joined #salt
22:08 huleboer joined #salt
22:10 peters-tx2 It got it to work
22:10 peters-tx2 But just as a complete debug test.
22:10 Hex_Offender joined #salt
22:11 peters-tx2 fllr, https://gist.github.com/PeterS242/0d1544e181140989b7ce
22:12 peters-tx2 I created a state for just this one box, and it applies two states, one is green and Result: True, while this second one is red and Result: False
22:13 aquassaut joined #salt
22:13 bash1245_ joined #salt
22:14 hasues joined #salt
22:14 hasues left #salt
22:16 peters-tx2 The "changed" is unneccessary.
22:17 fllr peters-tx2: did you run that as a custom state, or custom module?
22:17 peters-tx2 hmm
22:21 wincyj joined #salt
22:24 thayne joined #salt
22:25 peters-tx2 fllr, well, not working as I expected but most likely because I'm too newb; I tried calling it as a state, but it's complaining that it is unavailable
22:30 rokkanon joined #salt
22:33 diegows joined #salt
22:34 peters-tx2 fllr, lol well I actually can't get anywhere...  State 'deleteme.deploy' found in SLS 'test' is unavailable
22:35 beauzeaux joined #salt
22:35 peters-tx2 fllr, So while it looked like it was running as a state, it wasn't
22:35 peters-tx2 But I *can* call it as an execution module, however the result codes really don't seem to be the same in that case
22:38 peters-tx2 Ok now I'm getting somewhere
22:38 peters-tx2 durp
22:39 fllr peters-tx2: Lol. I'm telling you...
22:39 fllr Salt has some nice features, but this one has definitely been glossed over. Lol
22:39 peters-tx2 Ok I got it working.
22:40 peters-tx2 Do you have your code in _states
22:40 fllr peters-tx2: No, bro. _modules
22:40 peters-tx2 I got it working as a state; not a module.
22:41 fllr peters-tx2: Yeah... I need it as a module...
22:41 aphorise joined #salt
22:42 triplecorn joined #salt
22:44 peters-tx2 Are you sure that execution modules can go red on errors?
22:47 triplecorn joined #salt
22:48 fllr peters-tx2: technically, highstate is a module... =p
22:49 fllr peters-tx2: I think the problem is the 'highstate' outputter is an outputter built for the highstate module only... as opposed to being an outputter that displays data 'like the highstate module'...
22:50 peters-tx2 fllr, That sounds about right
22:50 JDiPierro joined #salt
22:51 peters-tx2 fllr, I'll fiddle with it myself
22:54 fllr peters-tx2: Cool, I'd appreciate that!
22:54 fllr This issue has really been killing me...
22:55 fllr You'd figure that they'd already figured out a way to make this thing generic enough for other people who want to output stuff similarly to the highstate module...
22:59 bhosmer joined #salt
23:06 DorfOnGolf joined #salt
23:10 SheetiS joined #salt
23:13 nich0s joined #salt
23:15 thayne joined #salt
23:17 Gareth left #salt
23:33 I3olle joined #salt
23:39 yomilk joined #salt
23:40 peters-tx2 fllr, Well, I got it to work :P
23:41 fllr peters-tx2: how?
23:41 peters-tx2 fllr, It is the format of the ret
23:41 peters-tx2 fllr, 1 sec
23:42 beneggett joined #salt
23:42 peters-tx2 fllr, This should get you a long ways:  ret = {'deleteme_|-ID_|-name_|-deploy': {'comment': 'This is the reason why it failed, fill this out.', 'name': 'deploy', 'start_time':'', 'result': False, 'duration':'', '__run_num__':2, 'changes': 'These are the changes.' } }
23:43 peters-tx2 That first bit before the comment you can change it up
23:43 peters-tx2 Replace ID with whatever, and name with whatever
23:43 peters-tx2 I'm not sure how much of the other stuff is required; probably not
23:43 peters-tx2 Like run_num, duration etc
23:44 fllr Oh... I have to explicitly pass in the id of my machine, then?
23:44 fllr Also... what is hat run_num nonsense?
23:44 peters-tx2 Looks that way
23:44 peters-tx2 The run_num is for the "large mass of states" to order their output correctly
23:45 peters-tx2 I just deleted it and it works OK
23:45 fllr Awesome! I'm gonna try that right now! :)
23:45 peters-tx2 The way I found it was to run state.highstate -l all   for a machine and carefully look at that result code
23:46 peters-tx2 I figured that was what highstate output was using
23:46 fllr Ooooooh! Clever...!
23:46 fllr I never even thought of doing that...
23:46 peters-tx2 fllr, Sorry it isn't documented, but the code states "The HighState Outputter is only meant to be used with the state.highstate function, or a function that returns highstate return data."
23:47 fllr peters-tx2: Yeaaaah... I don't see why, though...
23:48 peters-tx2 fllr, I don't think they expect execution modules to ever output highstate data
23:48 desposo joined #salt
23:48 peters-tx2 fllr, Sooner or later I'm sure better colorization of modules will come about :)
23:48 fllr That output looks incredibly readable, and you can easily switch between full details, and short versions... I'm surprised it's not written in a more generic way... :p
23:49 fllr peters-tx2: Yep. I hope so... Lol
23:49 yomilk joined #salt
23:50 fllr peters-tx2: I've it working! :~~~~)
23:50 fllr Thanks, bro!
23:51 peters-tx2 Nice!
23:51 beneggett joined #salt

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