Perl 6 - the future is here, just unevenly distributed

IRC log for #salt, 2016-07-25

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

All times shown according to UTC.

Time Nick Message
00:03 ageorgop joined #salt
00:05 jtang joined #salt
00:10 nidr0x joined #salt
00:10 Bryson joined #salt
00:11 ninjada joined #salt
00:11 DEger joined #salt
00:12 nidr0x joined #salt
00:24 DEger joined #salt
00:25 flowstate joined #salt
00:30 DEger joined #salt
00:40 DEger joined #salt
00:51 ageorgop joined #salt
00:52 DEger joined #salt
00:57 raspy_ joined #salt
00:59 DEger joined #salt
01:02 edrocks joined #salt
01:05 Nahual joined #salt
01:05 bastiandg joined #salt
01:11 catpigger joined #salt
01:17 DEger joined #salt
01:25 flowstate joined #salt
01:28 flowstate joined #salt
01:29 spuder joined #salt
01:38 mohae joined #salt
01:43 amcorreia joined #salt
01:52 subsignal joined #salt
01:54 PerilousApricot joined #salt
01:57 ageorgop joined #salt
01:58 aswini joined #salt
02:15 bastiand1 joined #salt
02:22 DEger joined #salt
02:26 flowstate joined #salt
02:26 k_sze[work] joined #salt
02:27 ssplatt joined #salt
02:29 evle joined #salt
02:31 PerilousApricot joined #salt
02:32 ninjada joined #salt
02:41 ninjada joined #salt
02:50 ssplatt joined #salt
02:58 av_ joined #salt
03:01 spuder joined #salt
03:04 berto- joined #salt
03:05 edrocks joined #salt
03:07 DEger joined #salt
03:10 racooper joined #salt
03:14 rdas joined #salt
03:26 flowstate joined #salt
03:29 rem5 joined #salt
03:37 DEger_ joined #salt
03:41 rem5 joined #salt
03:44 ruxu joined #salt
03:45 spuder joined #salt
03:47 Jimlad joined #salt
04:05 Tyrm joined #salt
04:05 raspy_ joined #salt
04:10 raspy_ joined #salt
04:10 raspy_ hi all, is there a way to run a command that has spaces?
04:10 raspy_ "firewall-cmd --set-default-zone=trusted"
04:11 raspy_ it gives me this error " a function with whitespace is not supported, perhaps this is an argument that is missing a ":""
04:11 hemebond raspy_: Is that the name of your state?
04:12 hemebond Actually I don't know what module you're using or referring to here.
04:15 kshlm joined #salt
04:17 tristianc joined #salt
04:17 raspy_ hemebond: oh crap i think i saw what i did :D
04:17 raspy_ thx for pointing that out " - name "firewall-cmd --set-default-zone=trusted""
04:23 SpX joined #salt
04:24 kshlm joined #salt
04:25 flowstate joined #salt
04:28 aw110f_ joined #salt
04:32 Pulp joined #salt
04:37 keldwud joined #salt
04:37 keldwud joined #salt
04:39 ninjada joined #salt
04:49 theblazehen joined #salt
05:03 elias_ joined #salt
05:06 mailto1587 left #salt
05:07 zer0def joined #salt
05:08 edrocks joined #salt
05:09 PerilousApricot joined #salt
05:11 POJO joined #salt
05:24 aswini joined #salt
05:25 flowstate joined #salt
05:30 aswini joined #salt
05:46 jhauser joined #salt
05:47 kshlm joined #salt
05:58 SpX joined #salt
06:03 colttt joined #salt
06:05 ivanjaros joined #salt
06:11 ikarpov joined #salt
06:14 felskrone joined #salt
06:14 impi joined #salt
06:21 jtang joined #salt
06:24 k_sze[work] joined #salt
06:26 flowstate joined #salt
06:31 kawa2014 joined #salt
06:32 aswini joined #salt
06:36 Miouge joined #salt
06:39 aswini1 joined #salt
06:39 PerilousApricot joined #salt
06:43 Disorganized_ joined #salt
06:48 yuhlw_ joined #salt
06:58 felskrone joined #salt
07:01 sagerdearia joined #salt
07:05 ivanjaros joined #salt
07:06 Tyrm joined #salt
07:08 om joined #salt
07:11 edrocks joined #salt
07:13 AirOnSkin joined #salt
07:13 fracklen joined #salt
07:14 rdas joined #salt
07:16 fracklen joined #salt
07:28 infrmnt joined #salt
07:30 manji joined #salt
07:35 ninjada_ joined #salt
07:36 ruxu joined #salt
07:37 whitenoise joined #salt
07:40 PerilousApricot joined #salt
07:45 alvinstarr joined #salt
07:52 ub joined #salt
07:55 ub hi!
07:55 ub is it possible to access the job ID from inside a statefile?
07:55 kaushal_ joined #salt
07:55 ronnix joined #salt
08:01 sagerdearia joined #salt
08:09 cyraxjoe joined #salt
08:10 pcdummy ub: you mean from a the id of the current state, right?
08:10 ub yes
08:10 JohnnyRun joined #salt
08:11 ub i try to do something like salt.wait_for_event:
08:11 ub name: salt/netapi/hook/stagingtask/{{ jid }}/os_update
08:11 mortis im having a weird issue after i renamed a pillar for one of my minions
08:11 mortis it still sees the old name
08:11 mortis even after i clear the cache
08:11 mortis and restart
08:12 mortis both on minion and master
08:12 mortis greped through the master-cache and found the old name in a data.p file, delete that, restarted, still the same :)
08:14 pcdummy ub: not sure howto get the current JID
08:15 pcdummy mortis: minion name is in /etc/salt/minion_id on the minion if you mean that.
08:15 Hybrid joined #salt
08:16 ub pcdummy: any idea for a workaround?
08:17 ub pcdummy: is it possible to make a API request and use the json data in the state file?
08:17 Hybrid Hi
08:18 Hybrid Upragded recently to 2016.3.1 (Boron) and everything is so much slower with this release, I don't understand why
08:18 GreatSnoopy joined #salt
08:19 Hybrid I have a few very big state files and it already took quite a while with 2015 but really, it is easily 10 times slower now and I didn't change anything except upgrading both master and minions
08:19 Hybrid what could cause that ?
08:19 gmoro joined #salt
08:19 pcdummy ub: every state gets an kwarg "__pub_jid" may that helps
08:19 lero joined #salt
08:20 pcdummy ub: state = lxc/file and so on.
08:22 Hybrid even "salt-run jobs.active" is taking ages where it was extremely quick beforehand
08:23 pcdummy Hybrid: you have external job_cache and/or returner ?
08:24 Hybrid how would I know ?
08:24 pcdummy Hybrid: its a setting you have to enable, so surely no.
08:25 Antiarc joined #salt
08:25 Hybrid my setting is pretty out of the box ... juste configured my minions to tap on my master (I only have one) and I just added a few more thread capacity on the master but that's all
08:25 flowstate joined #salt
08:26 ribx joined #salt
08:26 Hybrid to me, it looks like some sort of DNS problem, everything regarding Salt has slow down since the upgrade to 2016.3.1 but I don't see why, as I said, except Salt itself, nothing has changed in my infrastructure
08:27 Hybrid and it was fast before
08:27 N-Mi joined #salt
08:31 ub pcdummy: i can access the __pub_jid only with python. right? do i have to write my own module?
08:33 impi joined #salt
08:35 s_kunk joined #salt
08:42 PerilousApricot joined #salt
08:43 ozux joined #salt
08:54 Rumbles joined #salt
09:01 ozux__ joined #salt
09:04 infrmnt joined #salt
09:05 Hybrid looking closely, seems I have a problem on the server side ... when calling my state files from the minions (with salt-call), it is working quite good, I would say at least as fast as it was before upgrading (the minion is 2016.3.1 too anyway)
09:05 Hybrid but calling the same statefile from the master is taking a lot lot lot longer
09:10 jtang joined #salt
09:12 ozux joined #salt
09:12 antpa joined #salt
09:14 edrocks joined #salt
09:15 ninjada joined #salt
09:19 fredvd joined #salt
09:24 flowstate joined #salt
09:24 N-Mi joined #salt
09:26 saltuser joined #salt
09:27 saltuser Hi, can anyone explain why group.present state output shows "passwd: x" ?
09:29 babilen saltuser: Could you paste the entire output to, say, http://refheap.com ?
09:29 babilen (or any of http://refheap.com, http://paste.debian.net, https://gist.github.com, http://sprunge.us, … really)
09:33 saltuser babilen: http://pastebin.com/ZqxB2v5X
09:34 babilen That's the single pastebin site I'm *very* reluctant to look at
09:35 saltuser While i'm creating another paste  - why is that? :)
09:35 krymzon joined #salt
09:36 babilen pastebin.com mangles input, takes forever to load, often makes us enter a CAPTCHA to see your paste and fills the screen with ads. It is ugly and crowded and I've seen many problems with whitespace ..
09:36 saltuser https://gist.github.com/anonymous/67598d6b05477ab606ec01bdbf45849e
09:36 babilen Could you also paste your state?
09:38 babilen Which version of salt is this?
09:38 saltuser Master is salt 2016.3.1 (Boron)
09:39 babilen And the minion?
09:39 saltuser Minions are 2015.5.0, 2015.8.1, 2016.3.1
09:42 babilen saltuser: Okay, that behaviour corresponds to the code. It calls https://github.com/saltstack/salt/blob/develop/salt/modules/groupadd.py#L82 which includes a 'passwd' entry
09:43 babilen https://docs.python.org/2/library/grp.html
09:43 babilen the (encrypted) group password; often empty
09:43 babilen There you go
09:43 babilen Does that cause problems?
09:44 Rumbles joined #salt
09:44 saltuser No, just made me wonder about the purpose
09:44 babilen "No particular purpose"
09:46 saltuser I'll consider myself enlightened again :)
09:49 ThomasJ joined #salt
09:52 Hybrid any idea on how I can fix this ? Pillar failed to render with the following messages
09:52 Hybrid I activated debug but I get no clue on where the problem is
09:53 Hybrid I juste get the concerned pillar file but not the precise error
09:53 babilen Take a look at the master log
09:53 Hybrid and what seems strange is that running my state file from the minion , it runs with no error, only when ran from the master, it fails
09:54 Hybrid I looked at it ... no clue on where the problem is
09:54 Hybrid it doesn't say much
09:56 babilen What does it say when you run "salt-run fileserver.update ; salt 'yourminion' saltutil.refresh_pillar" ?
09:56 babilen Anything special about your pillar setup?
09:57 Hybrid nothing special ... pretty "out of the box" master settings
09:57 Hybrid salt-run fileserver.update  => True
09:57 babilen So, what is the output of those commands and the logs? Could you paste that to one of http://refheap.com, http://paste.debian.net, https://gist.github.com, http://sprunge.us, … ?
09:57 saltuser Different topic, but does anybody know if there will be any official salt training courses in europe? They had a fiew in London/Paris but now it's just US.
09:58 Hybrid no error with saltutil.refresh_pillar too
09:58 Pulp joined #salt
09:58 babilen Hybrid: And nothing in the logs? When do you get the error?
09:59 Pulp joined #salt
09:59 Hybrid I get the error with :  salt "myminion" state.sls my.state from the master
10:00 babilen Okay, paste the output of that + master logs
10:00 Hybrid when doing : salt-call state.sls my.state from the minion then no error
10:00 babilen (and why not include the bit beforehand)
10:00 Pulp joined #salt
10:01 Hybrid this is the only error line I get (debug activated in the master log file) :
10:01 Hybrid 2016-07-25 12:00:53,611 [salt.utils.event ][DEBUG   ][4230] Sending event - data = {'fun_args': ['ldev.users'], 'jid': '20160725120053563538', 'return': ['Pillar failed to render with the following messages:', "Rendering SLS 'common' failed. Please see master log for details.", "Rendering SLS 'mysql' failed. Please see master log for details."], 'retcode': 5, 'success': True, 'cmd': '_return', '_stamp': '2016-07-25T10:00:53.610642', 'fun': 'state.sls', 'id':
10:02 Hybrid and this is only what I get in the console ...
10:02 Tyrm joined #salt
10:02 babilen Please include the master log
10:02 Hybrid really nothing else
10:03 babilen It would really help if you could provide more information .. even if there is nothing there ..
10:03 babilen You could also paste the pillar SLS files in question (common and mysql)
10:03 saltuser Set minion to debug mode as well and check that log?
10:03 babilen And how do you run the master in debug mode?
10:04 babilen (and so on)
10:05 Hybrid I run the master in debug mode by modifying the log_level directive in the master conf
10:06 babilen Is it problematic for some reason to just copy and paste a bunch of log messages to one of the pastebins along with the SLS files?
10:07 babilen I mean if there's nothing in there there's nothing in there, but it is quite hard to work with nothing :)
10:07 Hybrid no problem at all, this is just that while chatting, I was trying a few things but until now, still changed nothing
10:07 babilen Do you still get the error if you remove "common" and "mysql" pillar from that minion?
10:09 antpa joined #salt
10:10 Hybrid Can't really remove them (at least, not "common") because it is used by 99% of my state files
10:10 Hybrid I tried to disable them in the top pillar file but still, I get the error
10:11 saltuser salt 'minion' pillar.items   - is everything present that should be?
10:11 jtang joined #salt
10:14 babilen Hybrid: Stop your master, run "ps aux|grep salt" -- What's the output? Then run "salt-master -ldebug" and "salt 'yourminion' saltutil.refresh_pillar" and paste the output of that run along with the master log to one of the aforementioned pastebin sites. It would also help if you could paste the pillar files in question.
10:15 Hybrid I can't paste the pillar files ... it is full of confidential informations there
10:15 Hybrid I already checked the salt master process
10:16 Hybrid restared many times and checking each time it was cleanly stopped
10:16 Hybrid salt 'minion' pillar.items gives me plenty of informations but no error
10:17 Hybrid this is crazy
10:18 babilen And pillar.items includes the bits from common and mysql?
10:18 saltuser Do you include pillars from different files?
10:19 Hybrid yes, pillar.items include them all with no error
10:19 saltuser Including stuff from different files can create errors if the whitespaces are not aligned :)
10:19 Hybrid saltuser: no, I don't include pillar files from sls
10:20 saltuser Hybrid: have you checked minion debug logs? Just in case
10:21 babilen Does this happen if you run other states with state.sls ?
10:23 Hybrid saltuser: activated debug and checked the log on the minion ... nothing valuable there
10:25 Hybrid babilen: same problem when running other sls
10:25 flowstate joined #salt
10:25 babilen But not highstate?
10:25 Hybrid no, not highstate
10:25 babilen Okay, do you use environments?
10:25 Hybrid highstate is working
10:26 Hybrid none, I'm using base
10:26 babilen GitFS ?
10:26 Hybrid nope
10:27 Hybrid as I told you ... very simple setup
10:27 babilen Does it happen with this minion only or with other minions also?
10:27 Hybrid all of them
10:27 Hybrid and as I said, running these sls from the minion with salt-call works with no error
10:27 babilen And you really can't paste the output + logs of the master when you run it?
10:28 babilen "salt-run cache.clear_all" + "salt '*' saltutil.clear_cache" might be worth a try
10:30 jeddi joined #salt
10:31 antpa How can I upgrade salt-master if installed via the bootstrap method?
10:36 Hybrid concerning my other problem (slowliness of the salt-master), I see this in the minion's log in debug mode :
10:36 Hybrid 2016-07-25 12:35:53,723 [salt.transport.zeromq][DEBUG   ][1706] Initializing new AsyncZeroMQReqChannel for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506', 'aes')
10:36 Hybrid 2016-07-25 12:35:53,723 [salt.crypt       ][DEBUG   ][1706] Initializing new AsyncAuth for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506')
10:36 Hybrid 2016-07-25 12:35:53,806 [salt.transport.zeromq][DEBUG   ][1706] Initializing new AsyncZeroMQReqChannel for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506', 'aes')
10:36 Hybrid 2016-07-25 12:35:53,806 [salt.crypt       ][DEBUG   ][1706] Initializing new AsyncAuth for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506')
10:37 Hybrid I get thousands of them ... looks like it pass its time authenticating each other
10:37 Hybrid I can have dozens of these lines before getting a call to a file or something of interest
10:38 Hybrid is it a normal behaviour ?
10:39 Hybrid and calling my highstate from the master is painfully slow whereas it run within normal duration when called from the minion
10:40 Hybrid I don't understand and I'm getting really frustrated at upgrading salt ... 2015 was way quicker
10:41 saltuser Is your minion version newer than master?
10:41 babilen Minions are not newer, but all over the place
10:42 Hybrid salt-master 2016.3.1 (Boron)
10:42 Hybrid salt-minion 2016.3.1 (Boron)
10:42 babilen Which version of PyZMQ? (salt-minion --versions-report)
10:42 babilen (along with "salt-master --versions-report" on the master)
10:43 Hybrid on the minion : PyZMQ: 14.0.1
10:43 babilen Hybrid: And you are absolutely sure that you don't have any stray salt-master processes?
10:43 Hybrid on the master : PyZMQ: 14.4.0
10:43 Hybrid babilen: 100% sure
10:43 babilen How do you test?
10:43 Hybrid I even rebooted the whole server
10:44 yidhra joined #salt
10:44 saltuser Do you have a test environment?
10:44 Hybrid babilen: I stopped the service, checked with ps
10:44 PerilousApricot joined #salt
10:44 Hybrid saltuser: no
10:44 babilen Hybrid: Could you paste your /etc/salt/master file?
10:45 babilen And your /etc/salt/minion (on the minion)
10:46 Hybrid master (minues comments) : http://paste.debian.net/784858/
10:46 babilen I wonder if this is due to the ZeroMQ version mismatch
10:47 babilen Hybrid: Could you, please, set "pillar_safe_render_error: True" in your master configuration?
10:47 Hybrid in the minion, I only have the master server name, nothing else, everything is commented out
10:48 Hybrid commented I meant
10:48 Hybrid master and minions are installed from official saltstack repo
10:48 Hybrid and upgraded to latest release
10:49 Hybrid to give you an idea, my highstate was sended from the master at 12:24 ... it is still running 25 minutes later
10:52 Hybrid running the same state.highstate from a minion that match my first one (same configuration called) at 12:50 ... finished just right now at 12:52
10:52 babilen I would still, very much, appreciate it if you could stop the master, run "ps aux|grep salt", then "salt-master -ldebug", wait for it to settle down and then paste the entire output and (master log) when you run your state.sls bit
10:53 babilen And, I guess, I'll stick to 2015.8 for a little while longer
10:53 Hybrid I'm really trying not to convince myself to revert back to this release ....
10:55 Hybrid definitely, there is something with those lines :
10:55 Hybrid 2016-07-25 12:55:33,830 [salt.transport.zeromq][DEBUG   ][1706] Initializing new AsyncZeroMQReqChannel for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506', 'aes')
10:55 Hybrid 2016-07-25 12:55:33,830 [salt.crypt       ][DEBUG   ][1706] Initializing new AsyncAuth for ('/etc/salt/pki/minion', 'ldev-226.servicemagic.eu', 'tcp://10.70.0.10:4506')
10:56 Hybrid when doing salt-call from the minion (and activating the debug log), I see very few of them in the log where I get thousands of them in the same log when running the highstate from the master
10:57 Hybrid will try to disable security just to see if that changes anything
10:57 babilen Might be due to the ZeroMQ version mismatch
10:57 babilen "disable security" ?
10:57 Hybrid I must say that salt is eating 100% of one of my cores too on the minion ...
10:58 krymzon joined #salt
10:58 Hybrid well, as I said, I installed packaged release, this is no manual installation or installation from sources
10:58 babilen Yeah, absolutely. Just different platforms
10:59 Hybrid differences are there : master is running on Debian 8 where this peticular minion is running on Ubuntu 14.04 LTS
11:00 Hybrid but if we have to make sure that master and minion are running exactly same platform, we're having a problem then ...
11:00 babilen You shouldn't
11:00 Hybrid I shouldn't what ?
11:00 babilen It's working fine here with various different platforms (master is typically the newest)
11:03 Hybrid as for here
11:03 babilen Do you manage your minions with salt?
11:04 babilen (the processes that is)
11:04 Hybrid seemed to find something ... at least for the slowliness problem
11:04 Hybrid running a highstate to check
11:04 babilen mmh?
11:04 amcorreia joined #salt
11:04 babilen What did you dig up?
11:05 Hybrid in master conf, I found this directive : ping_on_rotate
11:06 babilen That's not in the configuration you pasted (http://paste.debian.net/784858/)
11:06 Hybrid it is set to False by default, I marked it as True and now, I don't have the exchange keys anymore and my highstate ran within normal duration (1-2 mn)
11:06 babilen And you typically want to have ping_on_rotate set
11:06 Hybrid no its not, I just chnaged it
11:06 Hybrid this is a default value
11:06 babilen aye
11:06 Hybrid it is "False" by default
11:07 Hybrid I set it to "True" and now its speedy again
11:07 Hybrid at least one point fixed
11:08 Hybrid but t didn't change anything concerning my pillar problem
11:09 babilen I still hope that looking at the debug output/logs would provide some information
11:09 Hybrid must leave now, getting back in 1h
11:09 Hybrid thanks for your help
11:13 jtang joined #salt
11:15 kawa2014 joined #salt
11:17 edrocks joined #salt
11:19 armin joined #salt
11:24 flowstate joined #salt
11:26 kshlm joined #salt
11:30 stack joined #salt
11:47 irctc531 joined #salt
11:48 irctc531 Does anyone here knows a salt package for Ubuntu webserver setup like does have nginx mariadb stuff?
11:48 lightus joined #salt
11:49 irctc531 Thanks in advance
11:50 babilen irctc531: I don't quite follow .. are you looking for salt states to configure mariadb and nginx?
11:51 fracklen irctc531: I would look for a formula for mariadb and nginx separately... And what would you have nginx do with mariadb?
11:52 irctc531 Yes exactly salt states, but already found them in github salt-formulas. I'm just looking for a package of states for a typical Ubuntu webserver
11:53 ninjada joined #salt
11:53 babilen What is a typical Ubuntu webserver?
11:54 irctc531 A typical ubuntu webserver? Well, it have nginx, mariadb, php-fpm, fail2ban, ssh, ftp
11:56 babilen You'll find formulas for all of them on github.com in saltstack-formulas
11:57 irctc531 I was planning to start my own web tech company and I just dont want to invest learning salt states as it was like rocket engineering for me
12:00 babilen irctc531: You can't really use salt if you don't want to learn it
12:02 irctc531 Well, I think it must study it now
12:03 irctc531 Is there any resource for Ubuntu server only?
12:03 irctc531 I seen some but its like all the OS in one PDF
12:04 Hybrid mastering salt IS complicated but not less than any orchestration plateform (I moved away from Puppet which, to me, was way less flexible)
12:05 Hybrid but on the other hand, it is very easy to start with salt ... just install the master on one server and the minion (client program) on each managed nodes and you're ready to go, you can at least, use the magical "cmd.run" command that lets you send commands to a whole group of nodes simultaneously
12:06 irctc531 Was that the best to practice?
12:06 babilen Hybrid: I honestly have no more clues .. it is baffling that a highstate from the master and minion works, while state.sls only works on the minion
12:07 babilen Hybrid: The best I can come up with is digging through the logs and searching for information. The pillar render error should have made it into the master log and I am surprised that it's not surfacing there
12:07 babilen You obviously checked that the minion and master processes are not being spawned twice or something crazy like that
12:07 flowstate joined #salt
12:08 babilen And just to make sure: The state you call with state.sls is part of state.highstate also, isn't it?
12:08 babilen Does it happen with "state.apply" too?
12:09 flowstate joined #salt
12:10 TooLmaN joined #salt
12:10 fracklen irctc531: As with most things in IT - Learning by doing... My best advice is to find a relevant task - E. g. set up a server with mariadb, nginx and php-fpm. Try to do this manually with saltstack. You'll learn a bunch from it.
12:12 impi joined #salt
12:14 jtang joined #salt
12:18 Hybrid babilen: yes, this sls (and the others) are part of the highstate
12:18 Hybrid and really, logs are silent for this error
12:19 Hybrid I found a directive which looks promising : pillar_safe_render_error
12:19 Hybrid but I tried to change that and it changed nothing
12:20 krymzon joined #salt
12:20 AndreasLutro you migh want to run saltutil.refresh_pillar to trigger pillar render errors
12:20 AndreasLutro otherwise it'll just read from cache
12:20 Hybrid AndreasLutro: I emptied the cache ... still same error
12:21 AndreasLutro also what did you set pillar_safe_render_error to
12:21 ninjada joined #salt
12:22 Hybrid as default, it set at True, I tried with False but that didn't change anything
12:22 AndreasLutro you set it in the master config and restarted the salt master, right?
12:23 Hybrid AndreasLutro: exactly
12:26 AndreasLutro I seem to recall pillar rendering errors being well hidden in 2015.8 but would have to test in production if I wanted to be sure. this is what I get if I have a pillar rendering error https://bpaste.net/show/3b252901e836
12:27 edrocks joined #salt
12:29 Hybrid *GASP* ... found it
12:29 babilen mmh?
12:30 Hybrid I had to issue saltutil.clear_cache + saltutil.refresh_pillar to make it work ... what I don't understand is that I already tried that after your recommendations this morning and it was still not working ...
12:30 Hybrid but now it works
12:30 Hybrid pretty disturbing
12:31 west575 joined #salt
12:31 babilen Are you sure that you ran saltutil.clear_cache and not just the (master side) salt-run clear cache?
12:31 Hybrid I'm quite sure of it (would need to parse my history to be sure) but I'm sure that I even manually removed /var/cache/salt/* on my test minion
12:32 babilen Okay
12:32 babilen Curious ..
12:32 Hybrid well .. sorry for the disturbance, anyway your all helped me fix my problem so many thanks to you
12:33 babilen That's okay .. no worries
12:34 babilen It was a bit of a weird problem
12:34 babilen (still is, really)
12:35 Hybrid I guess the upgrade to a newer release left some mess ... I don't know if the upgrade process cleans the cache ... maybe it should
12:35 kawa2014 joined #salt
12:37 gh34 joined #salt
12:38 stack joined #salt
12:38 stack joined #salt
12:40 west575 joined #salt
12:40 CeBe joined #salt
12:43 PerilousApricot joined #salt
12:44 edrocks joined #salt
12:45 numkem joined #salt
12:48 kshlm joined #salt
12:49 stack_ joined #salt
12:52 Hybrid well ... thought I was out of my slowliness problem ... might have called for victory a bit too early ... it is still sluggish at some point where it was a lot faster before
12:52 Hybrid but I'll see this later ... I have work to do and already spent too much time on this already
12:54 rem5 joined #salt
12:56 CeBe joined #salt
12:56 babilen Grrr .. formula upgrades really tend to break too much
12:57 edrocks joined #salt
13:02 stack joined #salt
13:02 TyrfingMjolnir joined #salt
13:09 subsignal joined #salt
13:12 mapu joined #salt
13:13 Miouge joined #salt
13:14 mohae_ joined #salt
13:14 FreeSpencer joined #salt
13:14 FreeSpencer joined #salt
13:15 ronnix joined #salt
13:15 jtang joined #salt
13:15 yuhlw_ Hi, I try to boot up on external_pillar. I do have configured in salt master config: "extension_modules: /srv/salt/module". Now I try to have a simple pillar script that do return {"test_key": "test_value"}. How can I achieve this ?
13:15 keekz joined #salt
13:15 west575_ joined #salt
13:16 racooper joined #salt
13:17 stack_ joined #salt
13:20 dyasny joined #salt
13:28 stack joined #salt
13:30 ribx joined #salt
13:31 sroegner joined #salt
13:31 west575 joined #salt
13:33 protoz joined #salt
13:35 tapoxi joined #salt
13:35 tapoxi hi everyone, how do I grant access to a specific runner under publisher_acl ?
13:35 perfectsine joined #salt
13:39 edrocks joined #salt
13:44 huyby joined #salt
13:45 rem5_ joined #salt
13:46 ssplatt joined #salt
13:46 POJO joined #salt
13:50 debian112 joined #salt
13:51 scoates joined #salt
13:52 Tanta joined #salt
13:53 ninjada joined #salt
14:02 jeddi joined #salt
14:06 syndikate joined #salt
14:08 Brew joined #salt
14:10 stooj joined #salt
14:11 antpa joined #salt
14:16 jtang joined #salt
14:16 rem5 joined #salt
14:19 ozux__ joined #salt
14:22 krymzon joined #salt
14:32 bowhunter joined #salt
14:33 impi joined #salt
14:33 JPT joined #salt
14:36 raspy_ joined #salt
14:36 west575 joined #salt
14:37 aswini joined #salt
14:42 Tyrm joined #salt
14:43 clalow joined #salt
14:49 Tyrm joined #salt
14:49 traph joined #salt
14:50 PoRNo-MoRoZ joined #salt
14:51 PoRNo-MoRoZ hello
14:51 PoRNo-MoRoZ i'm salt-newbie
14:51 PoRNo-MoRoZ how can i uncomment lines, with leading spaces ?
14:51 PoRNo-MoRoZ can't get it to work ><
14:52 Rumbles joined #salt
14:53 cmarzullo Generally I just manage the whole file. Otherwise you'll have some to do a regex.
14:53 cmarzullo https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.comment
14:54 PoRNo-MoRoZ mybashrc:
14:54 PoRNo-MoRoZ file.uncomment:
14:54 PoRNo-MoRoZ - name: /somepath/.bashrc
14:54 PoRNo-MoRoZ - regex: .alias .* --color=auto.
14:54 PoRNo-MoRoZ not working ><
14:54 PoRNo-MoRoZ yes i read docs about comment and uncomment
14:55 cmarzullo reminds me of the saying. I had a problem I solved with regex. Now I have two problems.
14:55 PoRNo-MoRoZ that's probably somehow linked with YAML striping spaces ?
14:55 PoRNo-MoRoZ :DDD
14:55 PoRNo-MoRoZ state working without leading spaces in file
14:55 PoRNo-MoRoZ and i tried to insert .* at start of regex
14:56 PoRNo-MoRoZ but i should recheck
14:56 PoRNo-MoRoZ sec
15:03 syndikate left #salt
15:03 PoRNo-MoRoZ not working ><
15:03 west575 joined #salt
15:06 hasues joined #salt
15:06 hasues left #salt
15:07 fracklen joined #salt
15:09 Andrew joined #salt
15:11 fracklen_ joined #salt
15:11 PoRNo-MoRoZ i'm runing 2016.3.0
15:12 Disorganized_ joined #salt
15:14 badon_ joined #salt
15:16 jtang joined #salt
15:21 _JZ_ joined #salt
15:23 krymzon joined #salt
15:25 cmarzullo sorry mate. Haven't delt with that state very much. I tend to just manage the whole file.
15:26 babilen PoRNo-MoRoZ: What is it doing?
15:26 babilen And you might want to enclose that regex in quotes
15:27 babilen → regex: '.alias .* --color=auto'
15:27 ssplatt PoRNo-MoRoZ: you may want https://docs.saltstack.com/en/latest/ref/states/all/salt.states.file.html#salt.states.file.line
15:27 PoRNo-MoRoZ it's have to uncomment multiple lines
15:27 PoRNo-MoRoZ pretty similar
15:27 PoRNo-MoRoZ basically i just want to make console on all my nodes COLORFUL
15:27 PoRNo-MoRoZ :)
15:28 PoRNo-MoRoZ https://github.com/saltstack/salt/issues/25345
15:28 saltstackbot [#25345][OPEN] file.uncomment: regex with multiple patterns ignores others if one is already uncommented | Hi,...
15:28 PoRNo-MoRoZ probably my case
15:28 PoRNo-MoRoZ i have uncommented line with same regex
15:28 PoRNo-MoRoZ just before commented ones
15:28 babilen Probably, yeah
15:28 PoRNo-MoRoZ i'll try to hack it with uniq regex
15:28 west575_ joined #salt
15:29 ssplatt PoRNo-MoRoZ: sounds like you may have more luck templating the entire file, as cmarzullo was refering to before
15:30 ivanjaros joined #salt
15:31 babilen These bugs really take the fun out of salt
15:31 cmarzullo If you template the bashrc file and wrap sections in feature toggles. So you pass pillar that will turn various features on and off.
15:32 PoRNo-MoRoZ that's sounds complex to me at my current level of salt knowledge :)
15:32 PoRNo-MoRoZ started just yesterday
15:32 PoRNo-MoRoZ i'll look at pillars
15:32 hasues joined #salt
15:33 cmarzullo Basically managing a single like becomes unwieldy pretty quick. Look to manage the whole file.
15:33 cmarzullo s/single like/single line/
15:33 hasues left #salt
15:33 phy1729 left #salt
15:37 PoRNo-MoRoZ file.managed with source ?
15:37 PoRNo-MoRoZ source stored on salt-master
15:38 PoRNo-MoRoZ probably i'll move to this one, using this for color schemes
15:38 PoRNo-MoRoZ *currently i'm already using it for color schemes
15:38 PoRNo-MoRoZ anyway thanks :)
15:38 babilen sorry for peskiness
15:39 Miouge joined #salt
15:40 fracklen joined #salt
15:40 ozux joined #salt
15:44 Rumbles joined #salt
15:52 Andrew joined #salt
15:52 scsinutz joined #salt
15:56 rem5 joined #salt
16:00 PoRNo-MoRoZ leading spaces ruining .uncomment aswell as .comment ><
16:00 PoRNo-MoRoZ looks like tomorrow i'm gonna used fully managed
16:00 PoRNo-MoRoZ and i'll take a look .line
16:01 PoRNo-MoRoZ *at
16:01 writtenoff joined #salt
16:05 hasues joined #salt
16:05 hasues left #salt
16:05 jhauser joined #salt
16:07 Miouge joined #salt
16:08 rem5 joined #salt
16:11 brotatochip joined #salt
16:11 bltmiller joined #salt
16:11 cro joined #salt
16:12 corichar joined #salt
16:17 writtenoff joined #salt
16:17 jtang joined #salt
16:21 PerilousApricot joined #salt
16:22 woodtablet joined #salt
16:23 brent_ joined #salt
16:24 woodtablet joined #salt
16:28 Miouge joined #salt
16:30 jgarr how can I see the actual command run on a minion from a module? eg puppet.run
16:30 scsinutz1 joined #salt
16:31 jgarr I didn't see the command by adding -l all to the salt command. I think I can see it on the message bus but am not sure how to watch it
16:31 babilen jgarr: You can follow the minion log or run "salt-call -ldebug state.highstate" (or state.sls or whatever you are after) on the minion directly
16:32 ldelossa joined #salt
16:32 ldelossa Hey guys, I'm trying to use pillar data to fill in teh cron module, but I'm having issues when my pillar data uses '*/1' for every one minute
16:32 ldelossa I get this error at high state
16:32 ldelossa Rendering SLS 'base:core.cron' failed: expected alphabetic or numeric character, but found '/'; line 4
16:32 Slimmons joined #salt
16:32 ldelossa But the value inthe yaml looks fine
16:34 jgarr babilen: thanks, that got me exactly what I needed
16:34 ub joined #salt
16:34 Slimmons I very frequently get this error, Salt request timed out. The master is not responding. If this error persists after verifying the master is up, worker_threads may need to be increased., but my worker_threads are fine, and I'm past zeromq 2.whatever  Any other ideas on how to troubleshoot?  I'm getting it mostly after I've run salt-master -l debug, and restarted salt-master
16:34 ldelossa http://hastebin.com/vapuxapene.sm
16:34 Slimmons but if i completely kill the master, and start it back, it works....sometimes
16:34 Slimmons but not most of the time
16:35 ldelossa I pasted a link with my pillar, the state, and the error message I receive. This looks like it hsould be straght forward.
16:37 whytewolf ldelossa: - minute: '{{ cron_details['minute'] }}'
16:37 whytewolf when using / you need to quote it
16:37 ldelossa That works!
16:38 ldelossa What's the rule of thumb, if there's a special character in the variable, quote it?
16:38 whytewolf when the docs show something quoted. it most likely is for a reason :P
16:38 brotatochip joined #salt
16:39 ldelossa The docs don't show how to use the modul with pillar data?
16:39 ldelossa OH
16:39 ldelossa I see now
16:39 ldelossa I'm placing the variable, inside rendered quotes.
16:39 ldelossa Okay that makes sense.
16:39 ldelossa Thanks lo
16:39 whytewolf yeap
16:40 ldelossa lol*
16:40 ldelossa Gracias.
16:41 scsinutz joined #salt
16:44 huyby joined #salt
16:53 Slimmons it looks like when I run salt -l debug, then I can't actually run any states, i'm just getting saltreqtimeouterrors....so i can't actually see what's wrong with my salt-api setup
16:57 edrocks joined #salt
16:58 cmarzullo do mean to say salt-api? or this just regular master <-> minion traffic?
17:02 Slimmons salt-api
17:03 Slimmons so, when i use the curl to run the salt-api, and my salt-master debug is running, I just get the timeout in the debug, like it can't connect to the minion
17:03 Slimmons which is correct, because when I start my master back, it can't run test.ping, b/c there's no connection
17:04 brotatochip joined #salt
17:04 guedressel joined #salt
17:05 Slimmons then if i stop the master several times, and run pkill salt-master, then restart it, it can connect again....then I run the curl again...it fails...i run debug, and it loses connection with the minion again
17:05 Slimmons rinse repeat
17:05 cmarzullo seems like you have some networking issues.
17:06 cmarzullo if the master can't salt.ping the minions it seems unlikely it's an issue with the salt-api
17:06 Slimmons agreed, but how do you think I could troubleshoot this?
17:06 Slimmons I can't watch the traffic via debug, b/c that breaks the connection to the minion
17:06 cmarzullo tcpdump the minions and master on the relavant pub sub ports. make sure traffice is getting to where it's supposed oto go.
17:07 cmarzullo perhaps your minions are trying to traverse with ipv6 or something.
17:08 Slimmons thanks, i'll try that
17:13 mpanetta_ joined #salt
17:14 Edgan joined #salt
17:15 ageorgop joined #salt
17:15 brotatochip joined #salt
17:16 bowhunter joined #salt
17:18 jtang joined #salt
17:18 rem5 joined #salt
17:20 amcorreia joined #salt
17:20 murrdoc joined #salt
17:27 fracklen joined #salt
17:28 Sokel Can a defaults.yaml that I'm importing have (for example) {{ grains['...'] }} in it or will it be ignored?
17:30 Pulp joined #salt
17:30 elias__ joined #salt
17:30 Miouge joined #salt
17:31 gladiatr joined #salt
17:32 scsinutz joined #salt
17:33 Edgan Sokel: https://paste.fedoraproject.org/395374/46799214/  I put yaml directly in the map.jinja, and this works.
17:35 Rumbles joined #salt
17:35 ssplatt joined #salt
17:35 GreatSnoopy joined #salt
17:37 Sokel Edgan: Interesting. Would {% import_yaml 'state/defaults.yaml' as state_defaults %} work similarly to what you have there then?
17:39 Edgan Sokel: I can only say try it
17:40 corichar joined #salt
17:42 murrdoc1 joined #salt
17:43 murrdoc1 joined #salt
17:43 ssplatt joined #salt
17:43 brotatochip joined #salt
17:44 murrdoc joined #salt
17:44 murrdoc joined #salt
17:46 TOoSmOotH joined #salt
17:47 TOoSmOotH joined #salt
17:48 s_kunk joined #salt
17:50 west575 joined #salt
17:51 scsinutz joined #salt
17:52 iggy You can not put jinja in imported yaml files
17:54 mrueg joined #salt
17:54 Sokel Figured as much then. How would I go about getting jinja into some pieces of it?
17:55 Sokel This is assuming I'd also be merging against pillar (which I don't know if pillar accepts jinja or not, I'm assuming it does not).
17:55 murrdoc joined #salt
17:55 flowstate joined #salt
17:56 iggy it's not much of a default if it requires jinja
17:56 iggy you should do whatever jinja stuff you want to do in the map.jinja
18:00 zmalone joined #salt
18:01 scsinutz joined #salt
18:02 huyby joined #salt
18:02 johnkeates joined #salt
18:02 teryx510 joined #salt
18:04 aw110f_ joined #salt
18:04 flowstate joined #salt
18:10 bltmiller joined #salt
18:12 brotatochip joined #salt
18:14 pcdummy iggy: nearly completed: https://github.com/pcdummy/saltstack-lxd-formula
18:15 pcdummy iggy: it has a bad requirement on a custom version of pylxd currently (with PRs upstream).
18:15 ajw0100 joined #salt
18:19 jtang joined #salt
18:20 cliluw joined #salt
18:21 PerilousApricot joined #salt
18:21 edrocks joined #salt
18:22 CeBe joined #salt
18:25 onlyanegg joined #salt
18:26 west575_ joined #salt
18:26 pcdummy How do i open a "salt://" file within my own execution module?
18:26 PerilousApricot joined #salt
18:27 iggy cp.* (cache_file, etc)
18:27 pcdummy iggy: nice thank you.
18:28 pcdummy no, VERY NICE! :)
18:30 teryx510 joined #salt
18:31 Rumbles joined #salt
18:32 scsinutz joined #salt
18:34 johnkeates joined #salt
18:41 Andrew joined #salt
18:42 mpanetta joined #salt
18:44 bowhunter joined #salt
18:47 stack_ joined #salt
18:48 robawt joined #salt
18:49 Tyrm_ joined #salt
18:50 Tyrm__ joined #salt
18:51 brotatochip joined #salt
18:55 huyby joined #salt
18:58 yidhra joined #salt
19:02 jtang joined #salt
19:02 CimmX joined #salt
19:04 CimmX hello all....     Has anyone seen an issue where salt minion is running on windows and is rebooted, but highstate does not run after the reboot even if the startup_states: highstate is set in the minion conf file?
19:05 ub joined #salt
19:06 ajw0100 joined #salt
19:06 JohnnyRun joined #salt
19:06 teryx510 joined #salt
19:09 CimmX it is
19:10 ub joined #salt
19:10 CimmX I set it in the profiles.conf file.... then create the hosts with salt-cloud.    Everything is working except that the rebooted windows will not run highstate.    If I run it manually from master...  then it works through the to next reboot....  then same issue.
19:11 CimmX With windows, the initial setup requires several reboots to be complete.   1 for a rename, 1 for a domain join, several for application loads.    I have this working using custom grains....    all except the rebooted windows host won't checkin with top.sls....
19:12 CimmX The docs say it should just work.....
19:13 cmarzullo is the key being accepted on the master from the new instance?
19:14 cmarzullo generated and deployed and all the things between?
19:14 cmarzullo oh nvm. more closer reading.
19:17 iggy try using the reactor to run the highstate (vs startup_states)
19:20 antpa joined #salt
19:20 bltmiller joined #salt
19:24 CeBe joined #salt
19:30 fracklen joined #salt
19:35 fracklen joined #salt
19:37 brotatochip joined #salt
19:38 fracklen joined #salt
19:40 MTecknology What would be the easiest way to verify the current salt config isn't going to cause issues if I try to restart the minion?
19:41 MTecknology if salt-call test.ping &>/dev/null; then   is the best I've come up with
19:41 rem5 joined #salt
19:42 MTecknology (not worried about states, just the minion configs and process
19:42 Rumbles joined #salt
19:42 iggy test the config before you apply it somewhere?
19:42 u joined #salt
19:44 MTecknology I'm not pushing out untested changes, but I want the salt minion to have a chance to flip me off if I tell it to restart and there's an old verison of salt and the new config would break the old salt-minion
19:44 CimmX Minion is working.... since I can apply the highstate from the master.    Minion key is accepted of course.     Doesn't seem to be a config issue since all sls files work when run manually
19:49 CimmX MKTech - You could start the service manually, then run a test.ping from master to verify before issuing reboots?
19:49 fracklen joined #salt
19:55 lero joined #salt
19:56 brotatochip joined #salt
19:59 Tyrm joined #salt
20:02 Edgan MTecknology: Do you run salt only when needed or scheduled?
20:02 MTecknology Edgan: when needed
20:02 Edgan MTecknology: monitoring software to check if the salt-minion is running and send an alert when it is not?
20:03 Tyrm_ joined #salt
20:04 mapu_ joined #salt
20:04 Edgan MTecknology: Otherwise you need a salt config parser/checker to pre-validate the configuration before restart
20:04 teryx510 joined #salt
20:05 Miouge joined #salt
20:08 MTecknology Edgan: seems like I should be able to somehow use salt-call as a config parser
20:10 Edgan MTecknology: You are going to write /etc/salt/minion.new, and then run salt-call --local against it. Then based on the results control if you update /etc/salt/minion?
20:11 Edgan MTecknology: You would only get a basic syntax checker, not a validator. It won't be connecting to the salt master, and checking if it can connect.
20:12 CeBe joined #salt
20:12 Edgan MTecknology: Which seems like you are doing your testing a little late. It sounds like you are building a safety net, because you didn't do enough testing up front.
20:13 MTecknology building a safety net in case something didn't show up during testing
20:13 Edgan MTecknology: If you really want to tie /etc/salt/minion to version of salt to avoid mismatches, there are easier ways to do that with templating.
20:14 MTecknology That's not what I want to do
20:14 MTecknology There are a lot of scenarios where it's possible to get a broken config on a client.
20:14 MTecknology all of them unlikely, but they exist
20:16 Edgan MTecknology: If you do mass break salt, it would seem to be a better use of time to built a way to fix the breakage then half ass a solution to try to avoid it.
20:16 Edgan MTecknology: The thought that comes to mind is salt-ssh to use your formulas to rewrite /etc/salt/minion and restart salt-minion
20:17 MTecknology and that sounds safer to you?
20:17 Edgan MTecknology: I don't see a complete solution to testing /etc/salt/minion on any one node at the last minute
20:18 MTecknology When did I say "last minute"?
20:18 Edgan MTecknology: You are saying a checker right before restarting salt-minion, on a per instance basis
20:18 Edgan MTecknology: or you trying to do a simulated salt CI?
20:19 brd joined #salt
20:19 MTecknology I guess we can call this a last minute "oh, shit... somebody broke something" check, ya
20:19 Tyrm joined #salt
20:19 Edgan No, it is a avoid someone breaking things at the last moment safety net
20:19 Edgan But I see no full proof way of doing it.
20:20 MTecknology Why didn't you say "I don't know" instead of trying to tell me I'm wrong?
20:20 Edgan It is like lint checkers vs syntax checkers vs actually valid code
20:20 Tyrm joined #salt
20:21 Edgan MTecknology: I am pointing out you are trying to down a clunky path, and there are various other solutions that seem to be more elegant to avoid your potential problem
20:21 lorengordon joined #salt
20:22 MTecknology Adding salt-ssh to manage configuration files is a horrible solution, especially considering I'd implement the exact same checks there and be back at the exact same question.
20:22 MTecknology just with a lot of extra steps
20:22 Edgan It is already broken at that point. So no point to the checks. I am not saying run it against everything at once, but only the broken ones after you broke them
20:23 MTecknology "after you broke them"
20:23 * MTecknology sighs
20:24 viq MTecknology: salt-call -c (something) --hard-crash would be my starting point. Maybe also keep a $( config_version - 1 ) somewhere to revert to if that fails?
20:24 Edgan MTecknology: yes, I am trying to tell you without a lot of work you always risk breaking your minions. You have limited time in your day. So if you don't already have a I broke it solution, write that first.
20:25 MTecknology viq: would test.ping be enough to validate it was able to load the config without troubles?
20:26 MTecknology wait... I know that answer
20:26 MTecknology viq: Thanks!!
20:26 * MTecknology hugs viq
20:26 viq MTecknology: I'm not sure, but I would assume that it loads config before doing any operations. Also, --hard-crash to make it extra trigger-happy
20:26 viq :D
20:26 armguy joined #salt
20:27 * MTecknology likes extra trigger-happy
20:27 viq I have no idea what that flag "really" does, but the description sounds promising
20:28 Tyrm_ joined #salt
20:28 Edgan Raise any original exception rather than exiting gracefully
20:28 armguy joined #salt
20:29 Edgan This is an important option, --retcode-passthrough
20:29 viq MTecknology: another "stupid idea" - have a mostly same config symlinked in a different place, with an extra file telling it to use a different user and port - try to start a full minion with that config, if that works, it should work for the "real" one
20:30 mpanetta_ joined #salt
20:30 Edgan viq: I thought of that. You would have to specify the port on the command line, and it is still an invalid test if the configuration defines a port number.
20:31 viq Aye, --retcode-passthrough sounds very useful for that as well
20:31 whytewolf personally I stopped using the minion config for anything other then master:
20:31 mpanetta_ joined #salt
20:32 Edgan whytewolf: what do you do instead?
20:32 * MTecknology is curious as well
20:32 whytewolf 99% of the minion config can be done through pillars
20:32 viq Edgan: there are two situations when port number could fail - outside of range of valid ports, and when something else is listening there. First one should be caught in pre-testing, second one needs debugging "always", though most likely means old minion didn't quit
20:32 whytewolf other then connecting information
20:32 POJO joined #salt
20:33 whytewolf https://docs.saltstack.com/en/latest/ref/modules/all/salt.modules.config.html#salt.modules.config.get
20:33 MTecknology The only minion config my servers get (at home) is /etc/salt/minion.d/master.conf  (and sometimes id.conf if relevant)
20:33 impi joined #salt
20:33 whytewolf most options use config.get
20:33 MTecknology master: salt.lustfield.net  <-- the entire file
20:33 whytewolf that is all i do as well MTecknology
20:34 whytewolf enough so that the minion can get infor from the master
20:34 MTecknology ah, ya, we do the same thing :)
20:35 whytewolf makes it much more difficult to get something wrong if it is something that won't change
20:35 MTecknology at work, they have a very old set of stuff and maintain a lot of old band-aids because of things that didn't exist in salt ... at the time
20:37 whytewolf humm. must be really old cause iirc the master/pillar/grain config set has been a part of salt for a long time.
20:37 felskrone joined #salt
20:39 whytewolf not sure when it went in though. i know i wasn't using salt that long before it did
20:39 MTecknology I still remember 0.15.x
20:40 MTecknology I didn't really find salt usable until it hit 0.17.2
20:40 bowhunter joined #salt
20:54 datamann joined #salt
20:54 flowstate joined #salt
20:55 Tyrm joined #salt
20:55 Tyrm joined #salt
20:57 s_kunk_ joined #salt
20:58 brotatochip joined #salt
20:58 lero joined #salt
20:59 s_kunk joined #salt
21:03 brd joined #salt
21:05 fridder joined #salt
21:08 clevodearia joined #salt
21:09 Rumbles joined #salt
21:14 bowhunter joined #salt
21:15 PerilousApricot Hello, I'm new to salt and I'm trying to read/understand the openssh-formula. If I read right, you can let the openssh state generate the server keypair, but it only saves the public (and not private) key back no the master (the miner_functions bit). Is that correct?
21:19 mTeK How can a get salt to put the correct network interface name in a file?
21:19 mTeK I'm using freebsd and the nework is called by driver.
21:20 mTeK So some are em0 some are fxp0 and a few others.
21:20 g3cko joined #salt
21:23 flowstate joined #salt
21:26 theblazehen joined #salt
21:27 onlyanegg joined #salt
21:35 flowstate joined #salt
21:38 onlyanegg joined #salt
21:40 Edgan Maybe grains['hwaddr_interfaces'] with a for loop and say if interface != lo
21:49 brotatochip joined #salt
21:49 whytewolf mTeK: well depends. Edgan's idea is  workable one. but if you just need the ip that is on a interface just use network.ip_addrs which lets you look up the ip directly with out knowing the interfaces
21:56 subsignal joined #salt
21:59 PerilousApricot I'm struggling with the same thing myself. We have an internal 10.x network and an external network. It's easy to detect the 10.x network since it's DHCP, but the external side needs a static config.
22:00 PerilousApricot I really don't want to, but I guess the only way to handle that would be to have a pillar (?) with "ext_if", "ext_ip" keys for each dual-homed node?
22:00 mTeK Need to call the network name info file to configure pf
22:02 PerilousApricot I'm coming from puppet-land, and it feels like the same sort of mind-bending I had to do when I took a functional programming class. The paradigm is real different, and I'm having to re-learn how to do everything
22:03 protoz joined #salt
22:03 flowstat_ joined #salt
22:05 Edgan mTeK: This is why I would disable the per driver/bus naming scheme if possible. In Linux it used to be eth[0-9], and became things like ens3. It makes some sense in the case of physical hardware, but it just makes life hardware in virtualization.
22:05 keldwud joined #salt
22:05 Edgan I mean life harder
22:09 flowstate joined #salt
22:09 flowstate joined #salt
22:14 subsignal joined #salt
22:16 stooj joined #salt
22:25 flowstate joined #salt
22:28 flowstate joined #salt
22:29 ssplatt joined #salt
22:37 ashmckenzie joined #salt
22:40 cyborg-one joined #salt
22:47 whitenoise joined #salt
22:53 antpa joined #salt
22:57 hemebond Well, you can find interfaces based on their subnet. Not sure if that'll work for you lot, but it's there.
22:58 ponyofdeath hi, anyone know why cmd.run is ignoring the -require states?
22:59 hemebond ponyofdeath: Error in your state probably.
22:59 ponyofdeath https://bpaste.net/show/7952ba5dbc15
23:00 hemebond Your require is at the same level as your cmd.run.
23:00 ponyofdeath https://bpaste.net/show/b936996ad97e
23:00 ponyofdeath tried that too
23:00 ponyofdeath sorry
23:01 rem5 joined #salt
23:01 hemebond No circular requires?
23:01 murrdoc salt-call -l warning state.show_lowstate  --out=json | jq -c '.[] | .[] | .__sls__' | uniq
23:01 murrdoc how to see states in your salt highstate , in order :)
23:02 ponyofdeath https://bpaste.net/show/09d5629c5b61
23:02 ponyofdeath hemebond: ^^
23:05 hemebond The init_jchem_psql_store requires that the service stop and start?
23:05 ponyofdeath yup
23:05 hemebond I didn't think require forced an order.
23:05 hemebond But then the state order will do that I suppose.
23:06 ponyofdeath hemebond: https://bpaste.net/show/bd65019f61a7 output of salt
23:06 hemebond ponyofdeath: Going to try out what murrdoc posted?
23:07 hemebond "Can not initialize while service is running"
23:07 ponyofdeath parse error: Invalid numeric literal at line 1, column 6
23:07 ponyofdeath hemebond: yup which means its not running the stop service pre-requisite
23:08 jtang joined #salt
23:08 hemebond "Service jchem-psql was killed"
23:08 Tyrm_ joined #salt
23:08 hemebond But then you start it immediately afterwards.
23:09 hemebond If you can't init while the service is running then you shouldn't be requiring start_jchem_psql.
23:09 hemebond Because that starts it.
23:09 antpa joined #salt
23:10 ZachLanich joined #salt
23:10 ponyofdeath hemebond: ahhh ok thats probably it
23:10 ponyofdeath hemebond: thanks!
23:10 hemebond 👍
23:11 ZachLanich Hey all, I'm having trouble with Vagrant & Ubuntu 16 w/ the Salt Bootstrap. I found a PR that supposedly fixed it in the "stable" branch, but I'm still getting failures. Is there any reason why Vagrant's Salt Boostrapper wouldn't be using the latest Stable Boostrap script?
23:11 ZachLanich PR: https://github.com/saltstack/salt-bootstrap/pull/852 - Merged here: https://github.com/saltstack/salt-bootstrap/commits/stable?page=3
23:11 saltstackbot [#852][MERGED] Ubuntu 16.04 LTS Xenial Support | What does this PR do?...
23:13 ZachLanich The PR correctly identifies Ubuntu 16.04 Xenial, so this error won't happen: Bootstrapping Salt... (this may take a while) mesg: ttyname failed
23:13 ninjada joined #salt
23:13 hemebond ZachLanich: Did you download the latest version of the bootstrap script and configure Vagrant to use it?
23:14 hemebond ZachLanich: https://www.vagrantup.com/docs/provisioning/salt.html
23:14 ninjada joined #salt
23:14 fracklen Any news on a release date for 2016.3.2?
23:15 whytewolf last i heard it was in QA. so should not be much longer
23:15 ZachLanich hemebond: I did try doing so, and it didn't seem to work, but I'm trying again as we speak to make sure I didn't screw something up.
23:24 flowstate joined #salt
23:29 antpa joined #salt
23:29 ZachLanich Still fails with new bootstrap! Here's a gist of the output: https://gist.github.com/zlanich/6e244160a5a0d780f2558553bf054758
23:30 ZachLanich hemebond: ^
23:30 onlyanegg joined #salt
23:31 jacksontj joined #salt
23:34 flowstate joined #salt
23:34 antpa joined #salt
23:36 armguy joined #salt
23:37 fracklen joined #salt
23:48 ZachLanich Trying a new vb box to see if that fixes it.
23:52 teryx510 joined #salt
23:55 ZachLanich It worked!!! The last box I used was bad for w/e reason.
23:55 ZachLanich Now if I could just get the virtualbox salt provisioner to work lol
23:57 west575 joined #salt

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