[Help] Installing k8s 1.4 on rancher 1.1.4

Is this possible ?

I can’t run k8s 1.2.4 as I am running centos 7.2 and it doesn’s seem to start kublet and proxy and I also need pet state.

How would I go about running from the latest images ?

Can I change the catalog that launches the environments ?

@andreimc in the kubernetes template there is a minimum_rancher_version check specified in the rancher-compose.yml file. You’d have to modify that and I’m not sure how successful you would be deploying k8s 1.4 on rancher 1.1.x. It’d be interesting to hear if you get that working.

this also might be helpful in getting it working: http://docs.rancher.com/rancher/v1.2/en/catalog/private-catalog/#rancher-catalog-templates

So what I had in mind was:

1 disable rancher certified

  1. add my fork of the catalog where I add version 12 of the kubernetes stack and I change the image to 1.4.6-rancher2

  2. change my environment from cattle to a kubernetes environment.

Things I am not sure about, would changing the environment use my catalog when launching k8s ?

it should be as easy as adding your own catalog with the modified k8s system template.

Oh cool :smiley:

Thanks, I will give that a go now

I have added my catalog, when I switch my environment from cattle to kubernetes it still uses the old version in the rancher catalog.

If I try to install it on cattle I get an error to do with the uuid …

If I disable the rancher catalog nothing happens when I switch environment from cattle to k8s …

So getting closer to a working 1.4 cluster

I just need to sort this out :slight_smile:

11/16/2016 2:37:06 PMerror: failed to run Kubelet: could not init cloud provider “rancher”: Could not create rancher client: Failed to find schema at [https://rancher.example.com/v2-beta]

I don’t have the v2-beta api :slight_smile:

Which version of k8s should I use that doesn’t assume the v2 api ?

1.2.4 doesnt work for me maybe 1.3 ?

@andreimc have you tried running rancher v1.2.0-pre3? That might be easier than cobbling together a solution on 1.1.4.

v1.2.0-pre3 doesn’t install for me, it fails to start the agent and the load balancer. Should I try upgrading from 1.1.4 ?

I gave it a shot and i get the following pre3

time="2016-11-16T10:14:40Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:40Z" level=info msg="Can not launch agent right now: exit status 1" component=service
time="2016-11-16T10:14:46Z" level=error msg="Failed to open project management: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:46Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:46Z" level=info msg="Can not launch agent right now: exit status 1" component=service
time="2016-11-16T10:14:51Z" level=error msg="Failed to open project management: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:51Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:51Z" level=info msg="Can not launch agent right now: exit status 1" component=service
time="2016-11-16T10:14:56Z" level=error msg="Failed to open project management: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:56Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:14:56Z" level=info msg="Can not launch agent right now: exit status 1" component=service
time="2016-11-16T10:15:02Z" level=error msg="Failed to open project management: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:15:02Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:15:02Z" level=info msg="Can not launch agent right now: exit status 1" component=service
time="2016-11-16T10:15:07Z" level=error msg="Failed to open project management: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:15:07Z" level=fatal msg="Failed to read project: Can not create a stack, check API key [B147675B81A13E5246AF] for [http://172.17.0.1:18080/v1/projects/1a7/schemas]"
time="2016-11-16T10:15:07Z" level=info msg="Can not launch agent right now: exit status 1" component=service

@aemneina ^^ - not sure which version actually works might try going pre4-rc

pre4-rc4 @aemneina

time="2016-11-16T10:20:33Z" level=info msg="Can not launch agent right now: Server not available at http://172.17.0.1:18080/ping:" component=service
time="2016-11-16T10:20:38Z" level=info msg="Waiting for server to be available" component=cert
time="2016-11-16T10:20:38Z" level=info msg="Can not launch agent right now: Server not available at http://172.17.0.1:18080/ping:" component=service

pre-rc5

stuck here:

time="2016-11-16T10:29:04Z" level=info msg="[4/10] [go-machine-service]: Starting "
time="2016-11-16T10:29:04Z" level=info msg="[4/10] [rancher-compose-executor]: Starting "
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting websocket-proxy : Service websocket-proxy must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=info msg="[4/10] [load-balancer]: Starting "
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting rancher-compose-executor : Service rancher-compose-executor must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: rancher-compose-executor : Service rancher-compose-executor must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: websocket-proxy : Service websocket-proxy must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting websocket-proxy-ssl : Service websocket-proxy-ssl must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting go-machine-service : Service go-machine-service must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: websocket-proxy-ssl : Service websocket-proxy-ssl must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=info msg="[4/10] [load-balancer-swarm]: Starting "
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting load-balancer-swarm : Service load-balancer-swarm must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=error msg="Failed Starting load-balancer : Service load-balancer must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: load-balancer : Service load-balancer must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: load-balancer-swarm : Service load-balancer-swarm must be state=active or inactive to upgrade, currently: state=updating-active"
time="2016-11-16T10:29:04Z" level=error msg="Failed to start: go-machine-service : Service go-machine-service must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=fatal msg="Service rancher-compose-executor must be state=active or inactive to upgrade, currently: state=upgrading"
time="2016-11-16T10:29:04Z" level=info msg="Can not launch agent right now: exit status 1" component=service

pre4-rc6

time="2016-11-16T10:32:20Z" level=info msg="Waiting for server to be available" component=cert
time="2016-11-16T10:32:20Z" level=info msg="Can not launch agent right now: Server not available at http://172.17.0.1:18080/ping:" component=service
time="2016-11-16T10:32:25Z" level=info msg="Waiting for server to be available" component=cert
time="2016-11-16T10:32:25Z" level=info msg="Can not launch agent right now: Server not available at http://172.17.0.1:18080/ping:" component=service

@andreimc are you reusing the hosts? Did you wipe /var/lib/rancher in between deployments? Also are you using docker inc’s docker? If so which version? Is this an HA deployment? Also are you reusing the DB in these instances?

@aemneina I reused the hosts, I wiped /var/lib/rancher in between the deployments but kept db. I have never been able to set up HA deployment on anything higher than 1.1.4.

wondering if a fresh db would help here, not sure if there’s any cruft interfering.

I have tried on a fresh db as well. It doesn’t seem to work.

Currently I got k8s 1.4.6 running on 1.2.0-pre3 (not ha version)

k8s 1.4.6 seems to work as far as I can tell :smiley:

awesome! you seem to be on the bleeding edge with this :smiley: