Hide Forgot
Description of problem: Tried out latest installer (from the master) and looks like the default ingress create the router in the pod network which was not the case before (atleast from 0.14.0 tag). This is now blocking the console pod to get route info. Version-Release number of selected component (if applicable): $ openshift-install version openshift-install unreleased-master-550-g507b62e7609fb54abfb4357395820b5fd8b6d635 $ oc adm release info --commits | grep ingress cluster-ingress-operator https://github.com/openshift/cluster-ingress-operator 51389c54efb260d7827ce4e39fc520c0d2b2b695 $ oc get endpoints router-internal-default NAME ENDPOINTS AGE router-internal-default 10.129.0.22:80,10.129.0.22:1936,10.129.0.22:443 105m $ oc get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE router-default-59956f7589-hnc5k 0/1 Pending 0 31m <none> <none> <none> router-default-59956f7589-zmg7k 1/1 Running 0 131m 10.129.0.22 test1-9wlwg-worker-0-c5dsx <none> $ ssh core.126.51 sudo netstat -ntpl (nothing on the host bind to route port) Active Internet connections (only servers) Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name tcp 0 0 127.0.0.1:10248 0.0.0.0:* LISTEN 2907/hyperkube tcp 0 0 192.168.126.51:9100 0.0.0.0:* LISTEN 7417/kube-rbac-prox tcp 0 0 127.0.0.1:9100 0.0.0.0:* LISTEN 7160/node_exporter tcp 0 0 0.0.0.0:111 0.0.0.0:* LISTEN 1/systemd tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 2935/sshd tcp 0 0 0.0.0.0:47223 0.0.0.0:* LISTEN 2996/rpc.statd tcp 0 0 192.168.126.51:10010 0.0.0.0:* LISTEN 2994/crio tcp6 0 0 :::44040 :::* LISTEN 2996/rpc.statd tcp6 0 0 :::10250 :::* LISTEN 2907/hyperkube tcp6 0 0 :::9101 :::* LISTEN 3447/openshift-sdn tcp6 0 0 :::111 :::* LISTEN 1/systemd tcp6 0 0 :::10256 :::* LISTEN 3447/openshift-sdn tcp6 0 0 :::22 :::* LISTEN 2935/sshd tcp6 0 0 :::9537 :::* LISTEN 2994/crio Steps to Reproduce: Get the latest installer from master and build for libvirt provider. Start installer with libvirt provider and wait till it is up. Actual results: router is bound to pod network not to host network Expected results: router should bound to host network like it do for 0.14.0 tag version. $ oc adm release info --commits | grep ingress cluster-ingress-operator https://github.com/openshift/cluster-ingress-operator e53dfea77b35656f105c41d5c1a3bcb2bc6fbcba $ oc get endpoints router-internal-default -n openshift-ingress NAME ENDPOINTS AGE router-internal-default 192.168.126.51:80,192.168.126.51:1936,192.168.126.51:443 44m
Spoke with Eric Paris about how to handle libvirt generally right now. We're to assign this to 4.1 but keep the priority and severity low to ensure the bug doesn't get tracked as a GA blocker. It's unlikely anyone on this team will be looking into anything related to libvirt for the release, but we're happy to provide guidance if someone else wants to contribute.
Hi Praveen, could you please help check if this works on the libvirt?
$ openshift-install version openshift-install unreleased-master-593-gd1d142ded769e05e6e87764484872b81311195c1 built from commit d1d142ded769e05e6e87764484872b81311195c1 $ oc adm release info --commits | grep ingres cluster-ingress-operator https://github.com/openshift/cluster-ingress-operator e49c483cea90d0360ce653afdc8104e145d67123 $ oc project openshift-ingress Now using project "openshift-ingress" on server "https://api.test1.tt.testing:6443". $ oc get ep NAME ENDPOINTS AGE router-internal-default 192.168.126.51:80,192.168.126.51:1936,192.168.126.51:443 81m This is now fixed and now route is on host network not on pod network.
Thank you Praveen, so I move it to verified.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2019:0758