Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Conversation

derekwaynecarr
Copy link
Member

This PR makes the following changes:

  • installs etcd on host to hold flannel config
  • install configure flannel on master and nodes
  • update salt to take a flag to not remove host etcd setup

Fixes #13920
Fixes #11119

Once this lands, I will back port it to 1.0.x release stream.

@derekwaynecarr
Copy link
Member Author

As for why I cannot use the kubelet provisioned etcd to hold flannel configuration, its a chicken/egg problem since I need to configure flannel before I install/run docker and/or kubelet.

@bprashanth
Copy link
Contributor

Fyi #13877
Not that it's going in soon, but sometime we will solve the chicked and egg problem for you. in fact flannel won't need access to etcd.

@derekwaynecarr
Copy link
Member Author

@bprashanth - I would like this to go in before #13877 because it's easier to backport to 1.0.x stream, works, and unblocks folks. I hope this change will make it easier to then have vagrant included in the changes for #13877

@k8s-github-robot k8s-github-robot added the size/L Denotes a PR that changes 100-499 lines, ignoring generated files. label Sep 15, 2015
@k8s-github-robot
Copy link

Labelling this PR as size/L

@k8s-bot
Copy link

k8s-bot commented Sep 15, 2015

GCE e2e build/test passed for commit 360e762.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Should this be in an else block? For the platforms that don't have a host etcd, we don't need this rule to execute.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

No.

Even though we run the default kubernetes etcd via a pod, we bind mount that directory in for logging, so we need to touch the files first in order to allow the mount to proceed.

See https://github.com/kubernetes/kubernetes/blob/master/cluster/saltbase/salt/etcd/etcd.manifest#L41

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Got it, thanks.

@tsl0922
Copy link

tsl0922 commented Sep 16, 2015

@derekwaynecarr I can confirm this pr fixes #11119 for me, everything works fine, except that there is a failed unit(docker-storage-setup.service) on both master and minion, log:

[root@kubernetes-minion-1 vagrant]# journalctl -ru docker-storage-setup.service
-- Logs begin at Tue 2015-02-24 15:27:58 UTC, end at Wed 2015-09-16 00:04:44 UTC. --
Sep 15 23:25:00 kubernetes-minion-1 systemd[1]: docker-storage-setup.service failed.
Sep 15 23:25:00 kubernetes-minion-1 systemd[1]: Unit docker-storage-setup.service entered failed state.
Sep 15 23:25:00 kubernetes-minion-1 systemd[1]: Failed to start Docker Storage Setup.
Sep 15 23:25:00 kubernetes-minion-1 systemd[1]: docker-storage-setup.service: main process exited, code=exited, status=1/FAILURE
Sep 15 23:25:00 kubernetes-minion-1 docker-storage-setup[3186]: There is not enough free space in volume group vg_vagrant to create data volume of size MIN_DATA_SIZE=2G.
Sep 15 23:25:00 kubernetes-minion-1 docker-storage-setup[3186]: Logical volume "docker-poolmeta" created
Sep 15 23:24:59 kubernetes-minion-1 docker-storage-setup[3186]: Rounding up size to full physical extent 64.00 MiB
Sep 15 23:24:59 kubernetes-minion-1 systemd[1]: Starting Docker Storage Setup...

@derekwaynecarr
Copy link
Member Author

@tsl0922 - the docker storage setup unit failure is fine, and results in using devicemapper. I will move to overlayfs when its stable. Thanks for testing the feature!

@roberthbailey roberthbailey added lgtm "Looks good to me", indicates that a PR is ready to be merged. ok-to-merge labels Sep 16, 2015
@roberthbailey
Copy link
Contributor

LGTM.

@k8s-github-robot
Copy link

@k8s-bot test this [submit-queue is verifying that this PR is safe to merge]

@k8s-bot
Copy link

k8s-bot commented Sep 16, 2015

GCE e2e build/test passed for commit 360e762.

@k8s-github-robot
Copy link

Automatic merge from SubmitQueue

k8s-github-robot pushed a commit that referenced this pull request Sep 16, 2015
@k8s-github-robot k8s-github-robot merged commit 3dcb75c into kubernetes:master Sep 16, 2015
mikedanese added a commit that referenced this pull request Sep 16, 2015
…f-#13986-upstream-release-1.0

Automated cherry pick of #13986
@ghost
Copy link

ghost commented Sep 17, 2015

Works for me, although still have issues accessing https://10.245.1.2/api/v1/proxy/namespaces/kube-system/services/kube-dns, but think it is related to something else.

@ikehz ikehz mentioned this pull request Oct 27, 2015
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
lgtm "Looks good to me", indicates that a PR is ready to be merged. size/L Denotes a PR that changes 100-499 lines, ignoring generated files.
Projects
None yet
Development

Successfully merging this pull request may close these issues.

9 participants