=== axw_ is now known as axw
=== frankban|afk is now known as frankban
=== frankban is now known as frankban|afk
=== frankban|afk is now known as frankban
=== ken is now known as Guest54146
=== Guest54146 is now known as Ravenpi
[14:52] <Ravenpi> Trying to set up IPv6 endpoints for our cloud.  Have assigned 2005::100 to eth0 on my Keystone.  When I try to enable IPv6, though, I get "Interface 'eth0' does not have a scope global non-temporary ipv6 address." Ideas?
[14:53] <Ravenpi> (2005::100 shows as globally scoped via ifconfig.)
=== frankban is now known as frankban|afk
[19:21] <derekcat> Hey Everyone, quick question about the percona-mysql charm - is there a Juju way to change my.cnf options for the cluster?  For example, we need to set: ft_min_word_len=2
[19:29] <jose-phillips> hi derekcat
[19:30] <jose-phillips> try juju perscona-mysql config
[19:31] <jose-phillips> sorry
[19:32] <jose-phillips> juju config mysql
[19:33] <derekcat> Hey jose-philips!  Ok, give me a second, trying to get back into the controller.  >_<  Juju is hanging on the controller for this model...
[19:35] <stokachu> ryebot: gun1x  is having issues connecting to grafana
[19:35] <stokachu> ryebot: says it's asking for a password
[19:35] <gun1x> not only grafana ... all the services from these links ask for user/pass: https://bpaste.net/show/a43461bbec45
[19:36] <rick_h> gun1x: yea, grafana needs a password. The readme has instructions for getting it or you can change it via config
[19:36] <rick_h> gun1x: assume you mean for the webui?
[19:37] <gun1x> rick_h: i mean all the links from the paste
[19:37] <gun1x> none work
[19:37] <gun1x> if you have a readme, please share it
[19:37] <gun1x> this happened after deploying kubernetes from canonical, using conjure-up
[19:37] <rick_h> gun1x: https://jujucharms.com/grafana/
[19:38] <rick_h> gun1x: yea, try changing the password on grafana: juju config grafana admin_password=testing
[19:38] <rick_h> gun1x: looking at the paste for the other links
[19:39] <gun1x> gunix@juju-k8s:~$  juju config grafana admin_password=testing
[19:39] <gun1x> ERROR application "grafana" not found (not found)
[19:39] <rick_h> gun1x: oh! ok so these are services running in the k8 cluster, not the juju charms themselves
[19:39] <gun1x> yes.
[19:40] <rick_h> tvansteenburgh: or kwmonroe any ideas on the default proxy setup that gun1x needs? ^
[19:42] <gun1x> found it ... gunix@juju-k8s:~$ cat .kube/config.conjure-canonical-kubern-04a
[19:44] <gun1x> well ... now that i get pass auth ... grafana is not working, influxdb is not working and heapster is not working. they are all returning errors. omg
[19:56] <gun1x> heapster returns 404 page not found
[19:56] <gun1x> grafana returns
[19:57] <gun1x> {{alert.title}}
[19:57] <gun1x> and influxdb returns:   "status": "Failure",
[19:57] <gun1x>   "message": "no endpoints available for service \"monitoring-influxdb\"",
[19:57] <gun1x>   "reason": "ServiceUnavailable",
[19:57] <gun1x>   "code": 503
[19:57] <gun1x> i am confused how this failed. it is recommended by kubernetes and it is promoted on the ubuntu website.
[20:27] <jose-phillips> exist a way
[20:27] <jose-phillips> to allow perform changes inside of the container?
[20:27] <jose-phillips> and dont get deleted after reboot
[20:27] <jose-phillips> ?
[20:30] <stokachu> gun1x: think you need to run kubectl proxy
[20:30] <stokachu> gun1x: before accing those addons
[20:31] <gun1x> stokachu: from what i see kubectl proxy is working. is there any command to prove this ?
[20:32] <stokachu> gun1x: not sure unfortunately
[20:32] <stokachu> ryebot: ^
[20:32] <gun1x> :)
[20:33] <stokachu> gun1x: trying to get you some additional resources
[20:34] <stokachu> Cynerva: gun1x having trouble getting to grafana and other addons after running kubectl proxy
[20:34] <stokachu> https://www.irccloud.com/pastebin/4CYh6F4F/
[20:35] <gun1x> yep, that's it. but the kubernates dashboard works
[20:35] <gun1x> and it's behind the same proxy
[20:35] <stokachu> so just grafana then?
[20:35] <Cynerva> do metrics appear in the dashboard? that's where i'm used to looking for them
[20:36] <Cynerva> never tried hitting the services directly
[20:36] <gun1x> Cynerva: you mean grafana metric in the kubernates dashboard?
[20:38] <Cynerva> gun1x: pod metrics and such that come from heapster
[20:40] <gun1x> Cynerva: under workloads/pods within kubernetes-dashboard i get the following fields when clicking on a pod: details, containers, conditions, created by
[20:40] <gun1x> the pod list has name, anmespace, node,s tatus, retstards, age
[20:41] <gun1x> if something else is supposed to be here, i don't know
[20:41] <gun1x> i will try to destroy this deployment and do conjure-up again, but this time with calico instead of flanel
[20:42] <Cynerva> gun1x: hard to describe but it's usually pretty prominent, graphs and such on the pod overview, so they're probably missing
[20:42] <gun1x> yea, no graphs here
[20:45] <Cynerva> gun1x: if you haven't redeployed already, can you check `kubectl get po --all-namespaces` and make sure there's a influxdb-grafana-* pod and that it's in a Running state?
[20:46] <Cynerva> i've got a deployment coming up so i can get familiar with how it works again xD
[20:46] <gun1x> Cynerva: https://bpaste.net/show/9f1b4091fe3b
[20:49] <Cynerva> okay, that looks normal
[20:54] <knobby> maybe check the logs from the pods?
[20:55] <gun1x> knobby: lol Can't access the Grafana dashboard. Error: Get http://admin:admin@localhost:3000/api/org: dial tcp: i/o timeout. Retrying after 5 seconds...
[20:56] <gun1x> that is not the password
[20:56] <gun1x> however this log has no timestamp
[20:56] <gun1x> nvm found timestamp button
[20:57] <gun1x> knobby:  logs are from 2017-12-04T20:55:21.375298866Z ... not it's 22:57, 2 hours later
[20:58] <knobby> I meant for the pod itself: kubectl logs monitoring-influxdb-grafana-v4-hmt8j for example
[20:58] <gun1x> knobby:  those are logs from the pod :)
[20:59] <knobby> so the logs just say "Error: Get ..."?
[21:02] <knobby> gun1x, ^
[21:03] <gun1x> wait a sec, i need to google commands because kubectl has no autocompletion
[21:05] <gun1x> knobby: https://bpaste.net/show/0775416c87b3
[21:05] <gun1x> this is really confusing. i have no ideea why it still tries with admin:admin
[21:07] <gun1x> i tried from another browser and i get 2017-12-04T21:06:40.259719634Z t=2017-12-04T21:06:40+0000 lvl=info msg="Request Completed" logger=context userId=0 orgId=1 uname= method=GET path=/api/v1/proxy/namespaces/kube-system/services/monitoring-grafana/public/img/fav32.png status=404 remote_addr=192.168.122.1 time_ms=2ns size=23
[21:07] <gun1x> but still {{alert.title}}
[21:07] <gun1x> without any grafana content
[21:09] <gun1x> maybe it's redirecting badly and i don't get some of the javascript or css content
[21:09] <knobby> gun1x: my logs for my cluster show the same admin:admin error, but futher down say "connected to the Grafana dashboard."
[21:09] <gun1x> i will create a pod with novnc inside the deployment, maybe that works
[21:10] <knobby> gun1x: is your influxdb log ok? mine is full of 204's
[21:11] <gun1x> knobby: you mean this?
[21:11] <gun1x> 2017-12-04T21:11:05.087279702Z [httpd] 10.1.93.0 - root [04/Dec/2017:21:11:05 +0000] "POST /write?consistency=&db=k8s&precision=&rp=default HTTP/1.1" 204 0 "-" "heapster/v1.4.3" a3b1eb77-d937-11e7-8190-000000000000 203
[21:11] <knobby> same as mine
[21:12] <gun1x> you just did a VM with ubuntu 16.04 and did conjure-up ?
[21:15] <knobby> gun1x: bare metal and MaaS, but yes
[21:17] <Cynerva> gun1x: okay, i get the same results you do when i try to hit the services via kubectl proxy - 404 from heapster, weird {{alert.title}} page from grafana, no endpoint found for influxdb
[21:20] <Cynerva> for grafana, this bit from the browser console is fun: Loading failed for the <script> with source "http://localhost:8001/api/v1/namespaces/kube-system/services/monitoring-grafana/proxy/api/v1/proxy/namespaces/kube-system/services/monitoring-grafana/public/app/boot.85c49108.js".
[21:22] <gun1x> Cynerva: ...
[21:26] <Cynerva> gun1x: i'm able to get metrics from heapster at least, with a little url exploration - can you try something like http://localhost:8001/api/v1/namespaces/kube-system/services/heapster/proxy/api/v1/model/metrics/ ?
[21:26] <gun1x> Cynerva: i will try to deploy this manually without juju
[21:26] <gun1x> or with another solution, maybe openshift, i don't know
[21:30] <Cynerva> okay
[21:35] <stokachu> heh