onyx 2016-04-01 | Slack Archive

Drew Verlee00:04:21

im trying to follow the engraving tutorial and hitting im hitting a snag. Im guessing the problem is close to aws. Probably todo with some credentials im not setting correctly. If i run engraver cluster provision --cluster-id dev --ansible="-vvvv” everything seems ok, untill this... fatal: []: UNREACHABLE! => {"changed": false, "msg": "SSH encountered an unknown error. The output was:\nOpe… and then my play recap:

PLAY RECAP *********************************************************************
 : ok=0    changed=0    unreachable=1    failed=0
 : ok=0    changed=0    unreachable=1    failed=0
 : ok=0    changed=0    unreachable=1    failed=0
localhost                  : ok=32   changed=1    unreachable=0    failed=0

Using /Users/drewverlee/personal/engravings/one/hello-world/ansible/ansible.cfg as config file
 [WARNING]: provided hosts list is empty, only localhost is available

I would have expected 1’s in the ok box if everything went smooth. MY guess is that i need to setup something like a key for aws to use. when i did aws configuration I gave it my secret key and access id as well as user: ubuntu and pointed it at the .pem file i download. Aws shows my cluster is provisioned. But when i try to get logs i dont get anything… currently its complaining about docker. Which ill be honest, i dont understand, possible because i haven’t wrapped my head around docker entirely and possible because im not sure how its being used here. I assumed docker came in at the end to create images and ship them off to dockerhub. Do i need to be in a VM (virtualbox) and interacting with the docker engine (e.g docker run) during this entire setup process. (e.g run engraver cluster new --provider aws --cluster-id dev from inside a container.

bridget01:04:03

A similar thing happened to me when I ran engraver, also.

bridget01:04:47

Then I ran engraver cluster provision --cluster-id dev a second time, and everything worked. I never figured out what the problem was.

michaeldrogalis02:04:45

I was never able to reproduce what you @bridget have seen 😕

michaeldrogalis02:04:08

@drewverlee: You don't need to be able to call Docker until you run deploy

michaeldrogalis02:04:52

The provision step doesn't require local Docker.

jeroenvandijk11:04:17

When high availability isn’t a major concern (e.g. in case of an adhoc load test), is there an alternative for Zookeeper? E.g. can i choose a masternode that does the coordination that Zookeeper normally does?

bridget12:04:24

Zookeeper is required - it's fundamental to the way Onyx works

bridget12:04:20

If you are doing local testing, there is an embedded Zookeeper instance that you can use by including :zookeeper/server? true in your config

jeroenvandijk12:04:34

@bridget Cool, thanks

Drew Verlee14:04:05

@bridget. Do i need anything besides the information aws configuration requires? * access id * secret key * pem file * name of key * user

bridget14:04:50

Nope. That should be it.

bridget14:04:46

All I did was run engraver cluster provision --cluster-id dev two times, and, voila, everything was provisioned

bridget14:04:03

Seriously, just that command two times in a row

bridget14:04:42

If you have doubts about your aws configuration, you can review it in ~/.engraver

Drew Verlee14:04:20

would it be ok to drop a snippet of what i get when i run engraver cluster provision —cluster-id dev ? Im not sure there isn’t anything sensitive in that.

bridget14:04:41

There will be ip addresses... I can't think of anything else that would be a problem

michaeldrogalis14:04:17

@drewverlee: Yeah, definitely let me know what you're seeing. I'd like to nip this one in the bud.

bridget15:04:52

If this is the same thing I was seeing, I can go back through the steps to reproduce and create an issue. I can def reproduce it

2016-04-01

Channels