Hello, We have a quick question about the managed fluentd that is installed on GKE clusters with logging enabled. This particular cluster was created a few months ago. The installed fluentd DaemonSet appears to use a stackdriver-logging-agent image from around April:
gcr.io/stackdriver-agents/stackdriver-logging-agent:0.2-1.5.30-1-k8s (April 4) However, on GCR there are a number of images that are newer than this one. The latest is 0.4-1.5.35-1-k8s from Aug 16. The fluentd DaemonSet in our cluster does not appear to have been modified or updated since the cluster was created. Our question is whether it is possible to upgrade the fluentd DaemonSet to use a newer image tag, and how to do so. The documentation here <https://kubernetes.io/docs/tasks/debug-application-cluster/logging-stackdriver/#configuring-stackdriver-logging-agents> suggests that we can't manually modify the DaemonSet through kubectl (it's "managed and supported through GKE") -- if so, we are somewhat surprised that the fluentd DaemonSet hasn't been upgraded without our intervention. For context: the reason we are interested in upgrading fluentd is that we are experiencing segfaults in the fluent-plugin-google-cloud when exporting to Stackdriver. We noticed several somewhat-related issues being fixed in ruby-grpc and ruby-protobuf in the meantime, so we would like to try out a newer fluentd image to see if that resolves the issue. Thanks for your help! - Yifeng -- You received this message because you are subscribed to the Google Groups "Kubernetes user discussion and Q&A" group. To unsubscribe from this group and stop receiving emails from it, send an email to kubernetes-users+unsubscr...@googlegroups.com. To post to this group, send email to kubernetes-users@googlegroups.com. Visit this group at https://groups.google.com/group/kubernetes-users. For more options, visit https://groups.google.com/d/optout.
