Hi Michael, Thanks for your response. Apologies for the delay, I was trying to deploy a new OOM instance but I have run into the same tiller issue. I am using the deploy script in the integration repo according to the instructions in this link; https://wiki.onap.org/display/DW/OOM+Component. It creates a cluster of 13 nodes. I am running this instance on integration-SB-06 and from the earlier results of kubectl (before tiller failed). The cluster is not saturated. The IP of the VMs in both cases is static and did not change. Yes the helm init –upgrade was run. The rancher_vm_entrypoint.sh script run by the deploy script in integration shows that the token was displayed first. The OOM install completed and showed passed robot test cases before tiller failed. I cannot run any kubectl commands, when I try to run these commands, I get the following error;
The connection to the server 10.12.5.195:8080 was refused - did you specify the right host or port? Yes in getting the system setup and hosts attached, the commands here https://git.onap.org/logging-analytics/tree/deploy/rancher/oom_rancher_setup.sh#n209 were run. I am currently trying to readd hosts according to this https://git.onap.org/logging-analytics/tree/deploy/rancher/oom_rancher_setup.sh#n162 This problem is still a blocking issue for my testing HPA integration testing. Thanks, Itohan From: Michael O'Brien [mailto:[email protected]] Sent: Wednesday, October 24, 2018 3:26 PM To: [email protected]; Ukponmwan, Itohan <[email protected]> Subject: RE: Issues with Tiller (Cannot Connect to Tiller) Itohan, Hi, the k8s cluster is sensitive to network lag or timeouts – we are working this out if the out of the box k8s config is sufficient for all cloud-native deployments. On some systems the cluster may degrade over a couple days to a month – I have systems with 1500+ k8s container restarts that are ok after 3 weeks but on some clouds the cluster degrades faster – again the tool needs to be optimized fully. How many servers in your cluster? I ask this because you either can run 1 co-located master/cluster or 3+ nodes for etcd stability – unless you need to test actual failover – likely you are running multiple hosts or you would need the –max-pods=500 additional kubelet commands fix to go over 110. Is your cluster saturated – try to run with under 70% ram allocation – for example onap currently needs 160g – so you would need 220+G for a system that can reschedule pods to the remaining nodes based on a node failure. I would recommend several kubectl commands to triage your system – but if tiller is down then you won’t be able to do these Start with your ~/.kube/config If you are using root as well as your ubuntu? Non-root user – you would need to add these in both of your profiles on the master. Before this though – verify that the correct IP is being used when you registered your system either using the script or the “add host” button in the rancher gui. Another thing – did the IP of your VM change (try to use static/EIPs for the master – or register a domain name against it. – if the IP changes then the .kube/config will need to be upgraded. A third thing - did you wait until the rancher server displayed a token instead of user/pass in the add-hosts – this is a bug in 1.6 which is fixed by waiting a couple minutes after rancher_server docker startup – like the script does Worst case – just re-add your host(s) by regenerating the docker command from the gui – like the script https://git.onap.org/logging-analytics/tree/deploy/rancher/oom_rancher_setup.sh#n162 Did you run one of the scripts like the following to get your system up and hosts attached? Once you added hosts above https://git.onap.org/logging-analytics/tree/deploy/rancher/oom_rancher_setup.sh#n209 I ask this – just to verify you upgraded the server side of tiller sudo helm init --upgrade When tiller gets back up…. Do a list of nodes – for the k8s system you should see https://wiki.onap.org/display/DW/Cloud+Native+Deployment#CloudNativeDeployment-Scriptedundercloud(Helm/Kubernetes/Docker)andONAPinstall-SingleVM-256G Something like kubectl get pods --all-namespaces kube-system heapster-76b8cd7b5-g7p6n 1/1 Running 0 8m kube-system kube-dns-5d7b4487c9-jjgvg 3/3 Running 0 8m kube-system kubernetes-dashboard-f9577fffd-qldrw 1/1 Running 0 8m kube-system monitoring-grafana-997796fcf-g6tr7 1/1 Running 0 8m kube-system monitoring-influxdb-56fdcd96b-x2kvd 1/1 Running 0 8m kube-system tiller-deploy-54bcc55dd5-756gn 1/1 Running 0 2m Check your nodes Kubectl top nodes From: [email protected]<mailto:[email protected]> <[email protected]<mailto:[email protected]>> On Behalf Of Ukponmwan, Itohan Sent: Wednesday, October 24, 2018 5:29 PM To: [email protected]<mailto:[email protected]> Subject: [onap-discuss] Issues with Tiller (Cannot Connect to Tiller) Hi All, I have OOM ONAP deployed. It was working but now I have issues with connecting to tiller using helm or kubectl commands. When I do a helm version, I see the following; Client: &version.Version{SemVer:"v2.9.1", GitCommit:"20adb27c7c5868466912eebdf6664e7390ebe710", GitTreeState:"clean"} Error: cannot connect to Tiller Does anyone know how to fix this error? It is blocking my HPA integration testing. Thanks, Itohan “Amdocs’ email platform is based on a third-party, worldwide, cloud-based system. Any emails sent to Amdocs will be processed and stored using such system and are accessible by third party providers of such system on a limited basis. Your sending of emails to Amdocs evidences your consent to the use of such system and such processing, storing and access”. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#13306): https://lists.onap.org/g/onap-discuss/message/13306 Mute This Topic: https://lists.onap.org/mt/27623611/21656 Group Owner: [email protected] Unsubscribe: https://lists.onap.org/g/onap-discuss/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
