Akraino JIRA

Akraino JIRA https://jira.akraino.org This file is an XML representation of an issue en-us

9.4.5

940005

11-04-2023

[ICN-380] Install Istio for replicated control planes https://jira.akraino.org/browse/ICN-380 Integrated Cloud Native NFV <p>Install Istio for replicated control planes according to Istio's and Itohan's instructions:</p> <p><a href="https://istio.io/latest/docs/setup/install/multicluster/gateways/" class="external-link" target="_blank" rel="nofollow noopener">https://istio.io/latest/docs/setup/install/multicluster/gateways/</a></p> <p><a href="https://wiki.onap.org/display/DW/Istio+for+DCM" class="external-link" target="_blank" rel="nofollow noopener">https://wiki.onap.org/display/DW/Istio+for+DCM</a></p> <p> </p> <p>Will leverage Vagrant VMs and existing scripts from containerized KUD to significantly reduce time spent on this task and achieve ~2 story points.</p> ICN-380

Install Istio for replicated control planes

Story Medium Done Done Igor Duarte Cardoso Igor Duarte Cardoso Wed, 17 Jun 2020 19:36:24 +0000 Thu, 17 Dec 2020 00:54:39 +0000 Thu, 2 Jul 2020 22:36:18 +0000 0 1 <p>The specific cause of the issue above has not been found but I now know how this issue can happen vs. how not to have this issue. The two final tests were not modifying the coredns configmap as agressively as outlined in <a href="https://istio.io/latest/docs/setup/install/multicluster/gateways/" class="external-link" target="_blank" rel="nofollow noopener">https://istio.io/latest/docs/setup/install/multicluster/gateways/</a> and deploying k8s in a different way. </p> <p>I have been deploying KUD k8s on VMs using the KUD job, as follows:</p> <blockquote><p><tt>kubectl create <del>f </del></tt><br/> <tt>apiVersion: batch/v1</tt><br/> <tt>kind: Job</tt><br/> <tt>metadata:</tt><br/> <tt> name: kud-$CLUSTER_NAME</tt><br/> <tt>spec:</tt><br/> <tt> template:</tt><br/> <tt> spec:</tt><br/> <tt> hostNetwork: true</tt><br/> <tt> containers:</tt><br/> <tt> - name: kud</tt><br/> <tt> image: github.com/onap/multicloud-k8s:latest</tt><br/> <tt> imagePullPolicy: IfNotPresent</tt><br/> <tt> volumeMounts:</tt><br/> <tt> - name: multi-cluster</tt><br/> <tt> mountPath: /opt/kud/multi-cluster</tt><br/> <tt> - name: secret-volume</tt><br/> <tt> mountPath: "/.ssh"</tt><br/> <tt> command: <span class="error">["/bin/sh","-c"]</span></tt><br/> <tt> args: <span class="error">["cp -r /.ssh /root/; chmod -R 600 /root/.ssh; ./installer --cluster $CLUSTER_NAME --plugins onap4k8s"]</span></tt><br/> <tt> securityContext:</tt><br/> <tt> privileged: true</tt><br/> <tt> volumes:</tt><br/> <tt> - name: multi-cluster</tt><br/> <tt> hostPath:</tt><br/> <tt> path: /opt/kud/multi-cluster</tt><br/> <tt> - name: secret-volume</tt><br/> <tt> secret:</tt><br/> <tt> secretName: ssh-key-secret</tt><br/> <tt> restartPolicy: Never</tt><br/> <tt> backoffLimit: 0</tt></p></blockquote> <p>If, instead, I use the same exact setup and VMs but go into the VMs and manually install KUD there using:</p> <blockquote><p><tt>kud/hosting_providers/baremetal/aio.sh</tt></p></blockquote> <p>Then Istio will install correctly and I am able to test network connectivity between the 2 clusters.<br/> It is possible that some some KUD/ICN plugins are interfering with Istio because I have also disabled them when installing KUD manually via the aio.sh script. The reason I did it was the current failure of QAT plugin to install.<br/> To confirm, I have reproduced failure on KUD job 2 extra times and success on KUD aio.sh 2 extra times.<br/> The extra struggles around Istio certainly provided a better understanding of Istio than if no issues had been faced.<br/> Closing the story now.</p> <p>Still blocked.</p> <p>Additionally, attempted to start with a fresh deployment and now am blocked by a totally new issue that wasn't an issue 1-2 weeks ago:</p> <p> </p> <p>New pastebin of new issue: <a href="https://pastebin.com/JYwsCYjY" class="external-link" target="_blank" rel="nofollow noopener">https://pastebin.com/JYwsCYjY</a></p> <p>It seems like it's expecting a proxy but none are set (unless new code is setting them) and none are needed.</p> <p>Blocked due to lacking connectivity between istio clusters:</p> <p> </p> <p><a href="https://istio.io/latest/docs/setup/install/multicluster/gateways/" class="external-link" target="_blank" rel="nofollow noopener">https://istio.io/latest/docs/setup/install/multicluster/gateways/</a></p> <p> </p> <p>Can’t get traffic to go from one cluster to the other:</p> <p>$ kubectl exec --context=$CTX_CLUSTER1 $SLEEP_POD -n foo -c sleep – curl -I httpbin.bar.global:8000/headers</p> <p> </p> <p>The command is from the link above. That curl fails with a connection refused.</p> <p> </p> <p>The best I can find in the logs so far is this from the istio-proxy container of the $SLEEP_POD pod:</p> <p>transport: Error while dialing dial tcp: lookup istiod.istio-system.svc on 10.244.0.3:53: no such host</p> <p> </p> <p>A full log of that same istio-proxy container is here: <a href="https://pastebin.com/g1sDSxMG" class="external-link" target="_blank" rel="nofollow noopener">https://pastebin.com/g1sDSxMG</a></p> Relates ICN-381 ICN-376 Development Epic Link ICN-214 Rank 0|i004b0: Sprint ICN Sprint 20 ICN Sprint 21 Story Points 12.0