#deis

/

      • matthewwithanm has quit
      • c4milo has quit
      • c4milo joined the channel
      • c4milo has quit
      • viglesias has quit
      • kfarrell has quit
      • c4milo joined the channel
      • c4milo has quit
      • aledbf has quit
      • guidob joined the channel
      • c4milo joined the channel
      • c4milo has quit
      • gildegoma joined the channel
      • mil0 joined the channel
      • c4milo joined the channel
      • c4milo has quit
      • gildegoma has quit
      • mil0 has quit
      • mil0 joined the channel
      • gildegoma joined the channel
      • zz_aliem is now known as aliem
      • c4milo joined the channel
      • c4milo has quit
      • gamedna has quit
      • c4milo joined the channel
      • c4milo has quit
      • aledbf joined the channel
      • c4milo joined the channel
      • c4milo has quit
      • mil0 has quit
      • mil0 joined the channel
      • gildegoma has quit
      • mil0 has quit
      • mil0 joined the channel
      • mil0 has quit
      • mil0 joined the channel
      • [o__o] joined the channel
      • c4milo joined the channel
      • StDiluted has quit
      • StDiluted joined the channel
      • StDiluted has quit
      • zz_paulczar is now known as paulczar
      • StDiluted joined the channel
      • dscastro joined the channel
      • gamedna joined the channel
      • jchauncey joined the channel
      • aledbf has quit
      • aledbf joined the channel
      • kfarrell joined the channel
      • fozz
        I had some questions about deis usage
      • 1) How do you add a node to an existing cluster?
      • 2) how do I handle node configuration through etcd with deis?
      • 3) how do I manage service discovery and container monitoring?
      • Im trying to figure out how we could convert our current coreos/docker/fleet/systemd use case over to deis
      • bacongobbler
        fozz: 1) and 2) boot up a coreos node with the same DiscoveryURL. It'll be added to the cluster.
      • fozz
        will deis pick it up automatically and provision it or does deis need to do anything to new nodes at all?
      • mil0 has quit
      • bacongobbler
        fozz: it's all CoreOS, nothing Deis-specific
      • fozz
        oh snap
      • so our current configuration plan is to use systemd files to pull data out of etcd and into env variables, whats the suggest way with deis for node configuration?
      • bacongobbler
        so all you have to do is boot up a CoreOS node that has the same discovery URL set up as per CoreOS's documentation and it'll be added to the cluster, which the fleet scheduler will then handle
      • fozz
        perfect
      • bacongobbler
        fozz: can you give me a use case for "node configuration"? You mean like adding monitoring or container logging?
      • fozz
        I mean passing down postgres db url and credentials
      • for example
      • mil0 joined the channel
      • StDiluted
        $ deis config:set DATABASE_URL=postgres://
      • fozz
        ooohhh right
      • of course
      • those env variables are available within all nodes right?
      • StDiluted
        they are in etcd so yeah, I’m assuming they are pushed down to the nodes
      • fozz
        kickass
      • deis-bot has quit
      • bacongobbler
        oh, for applications? You have to set up a psql database somewhere external (or in the cluster with a docker container), and then use `deis config:set DATABASE_URL=<url>` to get the credentials into your app
      • fozz
        so we were planning on using a helper app running alongside a node setting ttl on a keepalive key to indicate the container was up and that the LB should be kept updated. I suspect this is somewhat impossible without container linking
      • bacongobbler
        this is following 12 factor app best practices: http://12factor.net/
      • fozz
        ty bacongobbler
      • deis-bot joined the channel
      • I understand you guys are working on a service discovery piece atm
      • bacongobbler
        fozz: as an example, you *could* try and see if deis/database would work for you guys with a bit of customization and spawn that as a separate container
      • we already have a service discovery piece. Unless you're referring to gh#231
      • [o__o]
      • fozz
        ooohhh
      • deis-bot has quit
      • bacongobbler
        botbot it awesome :)
      • fozz
        whats the existing piece?
      • bacongobbler
        etcd!
      • fozz
        so what data is populated in etcd regarding running apps?
      • I suppose its whatever the deis client pulls out
      • is there an external API our automation could use?
      • bacongobbler
        right now we don't have any way to automatically provision services to apps at the moment. the service registry feature is on my plate for v1.1
      • fozz
        gotcha
      • we could stick in a man in the middle app for now
      • bacongobbler
        fozz: the only data that's populated in etcd for running apps is in /deis/services, which is only the ip address and the port number
      • fozz
        thats all we need for adding to a LB
      • is there a way to provide a specific container the host ip so it can access etcd?
      • usually systemd would provide this
      • bacongobbler
        e.g. `etcdctl get /deis/services/fuzzy-wuzzy/fuzzy-wuzzy_v2.web.1` would return something like "172.8.17.100:49152"
      • fozz
        perfect
      • bacongobbler
        fozz: take a look in /etc/environment. It's how we get the host IP in the container :) https://github.com/deis/deis/blob/master/contro...
      • deis-bot joined the channel
      • fozz
        well you guys have just about thought of everything
      • bacongobbler
        there will be two envvars named COREOS_PUBLIC_IPV4 and COREOS_PRIVATE_IPV4
      • fozz
        kickass
      • thank you sir, I think that fills all my...err..holes
      • I'm gonna go dump this into a wiki doc
      • bacongobbler
        you are quite welcome
      • deis-bot has quit
      • StDiluted
      • bacongobbler
        hehe, I read that article yesterday. good read
      • StDiluted
        totally funny
      • and so much of it is true
      • bacongobbler
        I think "There will always be Darkness" rang so damn true for me
      • StDiluted
        as a primarily ops person who does dev only to support ops work, that is my life.
      • and even the reason I am in here right now
      • that article doesnt even cover the toll on the physical body, or the mental trauma of the usual work environment
      • dscastro has quit
      • paulczar is now known as zz_paulczar
      • wow
      • my router’s getting hit by php exploit attempts minutes after bringing it up
      • May 01 17:15:05 ip-10-183-157-237 sh[3146]: 2014/05/01 17:15:05 [error] 54#0: *778 open() "/usr/share/nginx/html/cgi-bin/php" failed (2: No such file or directory), client: 10.65.25.96, server: _, request: "POST
      • /%63%67%69%2d%62%69%6e/%70%68%70?%2d%64+%61%6c%6c%6f%77%5f%75%72%6c%5f%69%6e%63%6c%75%64%65%3d%6f%6e+%2d%64+%73%61%66%65%5f%6d%6f%64%65%3d%6f%66%66+%2d%64+%73%75%68%6f%73%69%6e%2e%73%69%6d%75%6c%61%74%69%6f%6e%3d%6f%6e+%2d%64+%64%69%73%61%62%6c%65%5f%66%75%6e%63%74%69%6f%6e%73%3d%22%22+%2d%64+%6f%70%65%6e%5f%62%61%73%65%64%69%72%3d%6e%6f%6e%65+%2d%64+%61%75%74%6f%5f%70%72%65%70%65%6e%64%5f%66%69%6c%65%3d%70%68%7
      • 3%67%69%2e%66%6f%72%63%65%5f%72%65%64%69%72%65%63%74%3d%30+%2d%64+%63%67%69%2e%72%65%64%69%72%65%63%74%5f%73%74%61%74%75%73%5f%65%6e%76%3d%30+%2d%64+%61%75%74%6f%5f%70%72%65%70%65%6e%64%5f%66%69%6c%65%3d%70%68%70%3a%2f%2f%69%6e%70%75%74+%2d%6e HTTP/1.1", host: "-h"
      • mil0 has quit
      • Are there plans to give feedback on progress when pushing and pulling from registries?
      • bacongobbler
        StDiluted: feedback on which operations? As in some progress in the middle of "git push deis master"?
      • we could probably pipe the output of https://github.com/deis/deis/blob/master/builde... to STDOUT
      • StDiluted
        yeah, something like that.
      • this appears to be why my container is breaking:
      • May 01 17:36:11 ip-10-185-95-212 sh[3779]: 2014-05-01 17:36:11 10.185.95.212:37210 onside-gemstone[web.1]: 2014/05/01 17:36:10 exec: "start": executable file not found in $PATH
      • bacongobbler
        StDiluted: that's an issue that we're going to have to address in v0.9.0. Unfortunately, dockerfile apps won't work unless you add a start script to the container similar to this one: https://github.com/opdemand/example-dockerfile-...
      • this broke once we added the scheduler
      • StDiluted
        oh, ok cool
      • i dont really need to run a Dockerfile app, i was just trying to run a test
      • would i be better off testing out the ruby/sinatra app?
      • YES! app published and working
      • so happy right now i could cry
      • zz_paulczar is now known as paulczar
      • mboersma
        woohoo!
      • epcim joined the channel
      • fxdgear is now known as zz_fxdgear
      • bacongobbler
        StDiluted: the two apps that we have been frequently testing with has been both example-ruby-sinatra and example-dockerfile-python
      • StDiluted
        ok, yeah I got the ruby sinatra one working
      • that’s closer to our app
      • bacongobbler
        our integration test suite uses ruby-sinatra as well
      • aliem is now known as zz_aliem
      • zz_fxdgear is now known as fxdgear
      • StDiluted
        weird. All was working, and then things stopped working suddenly. Then deis-builder.service went into stop mode