Fluentd setup broken in ansible setup.

[mohammedzee1000]

Kubectl version: Client Version: version.Info{Major:"1", Minor:"2", GitVersion:"v1.2.0", GitCommit:"5e723f67f1e36d387a8a7faa6aa8a7f40cc9ca46", GitTreeState:"clean"} Server Version: version.Info{Major:"1", Minor:"2", GitVersion:"v1.2.0", GitCommit:"5e723f67f1e36d387a8a7faa6aa8a7f40cc9ca46", GitTreeState:"clean"}

Environment:

• Hardware configuration: Lenevo Thinkpad T450s
  • OS (e.g. from /etc/os-release): Only machines running the kubernetes (vagrant kvm vms) NAME=Fedora VERSION="24 (Cloud Edition)" ID=fedora VERSION_ID=24 PRETTY_NAME="Fedora 24 (Cloud Edition)" ANSI_COLOR="0;34" CPE_NAME="cpe:/o:fedoraproject:fedora:24" HOME_URL="https://fedoraproject.org/" BUG_REPORT_URL="https://bugzilla.redhat.com/" REDHAT_BUGZILLA_PRODUCT="Fedora" REDHAT_BUGZILLA_PRODUCT_VERSION=24 REDHAT_SUPPORT_PRODUCT="Fedora" REDHAT_SUPPORT_PRODUCT_VERSION=24 PRIVACY_POLICY_URL=https://fedoraproject.org/wiki/Legal:PrivacyPolicy VARIANT="Cloud Edition" VARIANT_ID=cloud

• Kernel (e.g. uname -a):

Linux localhost 4.5.5-300.fc24.x86_64 #1 SMP Thu May 19 13:05:32 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux

• Install tools: dnf, ansible ansible rpm version : 2.0.2.0-1.fc24

What happened:

...
TASK [node : Setting the kubelet_modified fact to true] ************************
ok: [192.168.121.128]
ok: [192.168.121.138]

TASK [node : Install fluentd pod into each node] *******************************
fatal: [192.168.121.128]: FAILED! => {"changed": false, "dest": "/etc/kubernetes/manifests", "failed": true, "gid": 0, "group": "root", "mode": "0755", "msg": "Request failed", "owner": "root", "response": "HTTP Error 404: Not Found", "secontext": "unconfined_u:object_r:etc_t:s0", "size": 4096, "state": "directory", "status_code": 404, "uid": 0, "url": "https://raw.githubusercontent.com/kubernetes/kubernetes/master/cluster/saltbase/salt/fluentd-es/fluentd-es.yaml"}
fatal: [192.168.121.138]: FAILED! => {"changed": false, "dest": "/etc/kubernetes/manifests", "failed": true, "gid": 0, "group": "root", "mode": "0755", "msg": "Request failed", "owner": "root", "response": "HTTP Error 404: Not Found", "secontext": "unconfined_u:object_r:etc_t:s0", "size": 4096, "state": "directory", "status_code": 404, "uid": 0, "url": "https://raw.githubusercontent.com/kubernetes/kubernetes/master/cluster/saltbase/salt/fluentd-es/fluentd-es.yaml"}
    to retry, use: --limit @/home/moahmed/Work/kubernetes/contrib/ansible/playbooks/deploy-cluster.retry

What you expected to happen:

The ansible should not have failed at this point.

How to reproduce it:

Run the ansible script with appropriate inventory and setting "cluster_logging: true" in inventory/group_vars/all.yml

Anything else do we need to know:

I found that fluend-es is no longer available in kubernetes repository, instead being replaced by fluentd-gcp, located at the below url:

https://github.com/kubernetes/kubernetes/blob/master/cluster/saltbase/salt/fluentd-gcp/fluentd-gcp.yaml

[lucas-cegatti]

I had the same problem, here are some additional information.

The fluentd-es was migrated to Daemon Set in this commit.

https://github.com/kubernetes/kubernetes/blob/master/cluster/addons/fluentd-elasticsearch/fluentd-es-ds.yaml

For now I changed the raw link to release-1.5. As DaemonSet propagates through nodes I think it won't be necessary to create this pod on all nodes anymore.

[dustymabe]

I hit this as well, using the patch from the PR allowed me to get around it.