Hi,
Could you provide a bit more info on your cluster branch (assume
master/Casablanca oom), undercloud versions (docker/kubernetes/helm/rancher) =
(17.03/1.8.10.2.8.2/1.6.14 or 1.6.18).
How you deployed the cluster and attached the hosts - manually or via a
script.
Post details before/after you deploy some pods across the cluster like a top
output (if one of the cluster is detached - it will show here)
And a snapshot of the default(k8s+rancher) and onap namespaces "kubectl get
pods -all-namespaces -o wide
ubuntu@ip-172-31-28-152:~$ kubectl top nodes
NAME CPU(cores) CPU%
MEMORY(bytes) MEMORY%
ip-172-31-28-145.us-west-1.compute.internal 3699m 92% 26034Mi
85%
ip-172-31-21-240.us-west-1.compute.internal 3741m 93% 3872Mi
12%
ip-172-31-16-85.us-west-1.compute.internal 3997m 99% 23160Mi
75%
ip-172-31-25-15.us-west-1.compute.internal 3998m 99% 27076Mi
88%
The error you are quoting with etcd is independent of deploying ONAP pods or
any pod - if the clean cluster is not healthy then nothing will deploy
including onap containers.
Provisioning a Kubernetes cluster and deploying ONAP itself are two distinct
issues that can be done independently - I would recommend a two phase approach
1. Bring up a single node - etcd will always work in this case - run parts
of ONAP - bring up AAI or Logging for example (under the 110 pod limit)
2. Provision a cluster - keep it clean - verify all kubernetes services are
up (either via the rancher, kubernetes gui or via kubectl) - again bring up a
single pod that will scale across the nodes (use the disable all yaml and apply
a -set <component>.enabled=true on either a helm install or helm upgrade. - use
anything that has a replicaSet greater than 1 or deploy the log pod which has a
DeamonSet (container/VM).
For your reference:
There are clusters currently running in windriver/tlab and Amazon ok as a
reference.
The procedure/scripts for an openstack cluster is on
http://onap.readthedocs.io/en/latest/submodules/oom.git/docs/oom_setup_kubernetes_rancher.html
The procedure/scripts for an Amazon AWS cluster is on
https://wiki.onap.org/display/DW/Cloud+Native+Deployment#CloudNativeDeployment-4NodeKubernetesClusteronAWS
thank you
/michael
From: [email protected] <[email protected]> On Behalf Of
Deepak S
Sent: Tuesday, July 3, 2018 8:29 PM
To: [email protected]
Subject: [onap-discuss] Getting cluster unavailable error if OOM is deployed in
more than 1 node. <eom>
Hi All,
I am getting below error if I deploy OOM in more than 1 node
Error from server: client: etcd cluster is unavailable or misconfigured; error
#0: client: etcd member http://etcd.kubernetes.rancher.internal:2379 has no
leader
Thanks
Deepak
This message and the information contained herein is proprietary and
confidential and subject to the Amdocs policy statement,
you may review at https://www.amdocs.com/about/email-disclaimer
<https://www.amdocs.com/about/email-disclaimer>
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#10845): https://lists.onap.org/g/onap-discuss/message/10845
Mute This Topic: https://lists.onap.org/mt/23035460/21656
Group Owner: [email protected]
Unsubscribe: https://lists.onap.org/g/onap-discuss/unsub
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-
