Goal: I had deployed opensource istio 1.8.0 on GCP GKE Cluster which is also registered in as Anthos Cluster. Also this cluster is having Anthos ASM 1.8.1 pods (istioid-asm pods) deployed. Both the ASM and opensource Istio are deployed in istio-system namespace.
I have deployed opensource istio with the goal to achieve distributed deployment of an application on kubernetes clusters on Anthos GKE and Anthos AWS Cluster as defined in below article.
Hybrid-Multicluster-Service-Mesh on Anthos GKE and Anthos AWS Cluster
I am facing issues with the istio-ingressgateway and istio-egressgateway pods that are deployed on the Anthos GKE Cluster.
1) The istio-egressgateway is failing with Pod errors: CrashLoopBackOff and Does not have minimum availability errors failing with exit code 255 as shown in below screenshots.
After checking logs for the pods have found the error as :
**sidecarToTelemetrySessionAffinity: true]: 1 error occurred: * trustDomain: empty domain name not allowed**
I have tried to increase the cpu and resource limits for these pods and also mentioned the trustdomain as cluster.local in the deployment file but still facing the same issue with the istio-egressgateway pod.
2) The istio-ingressgateway is also failed with Pod errors: CrashLoopBackOff error.
Also the istio-ingressgateway pods are getting failed with exit code: 255