hello yes.. i have a external balancer in front of my masters for HA as doc says.
i don’t have any balancer in front of my etcd servers for masters connection, it’s not necessary right? masters will try all etcd availables it one is down right? i don’t know why but none of my masters were able to connect to the second etcd instance, but using telnet from their shell worked .. so it was not a net o fw issue.. best regards. > El 13 jun 2016, a las 17:53, Clayton Coleman <[email protected]> escribió: > > I have not seen that particular issue. Do you have a load balancer in > between your masters and etcd? > > On Fri, Jun 10, 2016 at 5:55 AM, Julio Saura <[email protected]> wrote: >> hello >> >> i have an origin 3.1 installation working cool so far >> >> today one of my etcd nodes ( 1 of 2 ) crashed and i started having problems.. >> >> i noticed on one of my master nodes that it was not able to connect to >> second etcd server and that the etcd server was not able to promote as >> leader.. >> >> >> un 10 11:09:55 openshift-balancer02 etcd[47218]: 12c8a31c8fcae0d4 is >> starting a new election at term 10048 >> jun 10 11:09:55 openshift-balancer02 etcd[47218]: 12c8a31c8fcae0d4 became >> candidate at term 10049 >> jun 10 11:09:55 openshift-balancer02 etcd[47218]: 12c8a31c8fcae0d4 received >> vote from 12c8a31c8fcae0d4 at term 10049 >> jun 10 11:09:55 openshift-balancer02 etcd[47218]: 12c8a31c8fcae0d4 [logterm: >> 8, index: 4600461] sent vote request to bf80ee3a26e8772c at term 10049 >> jun 10 11:09:56 openshift-balancer02 etcd[47218]: got unexpected response >> error (etcdserver: request timed out) >> >> my masters logged that they were not able to connect to the etcd >> >> er.go:218] unexpected ListAndWatch error: pkg/storage/cacher.go:161: Failed >> to list *extensions.Job: error #0: dial tcp X.X.X.X:2379: connection refused >> >> so i tried a simple test, just telnet from masters to the etcd node port .. >> >> [root@openshift-master01 log]# telnet X.X.X.X 2379 >> Trying X.X.X.X... >> Connected to X.X.X.X. >> Escape character is '^]’ >> >> so i was able to connect from masters. >> >> i was not able to recover my oc masters until the first etcd node rebooted >> .. so it seems my etcd “cluster” is not working without the first node .. >> >> any clue? >> >> thanks >> >> >> _______________________________________________ >> users mailing list >> [email protected] >> http://lists.openshift.redhat.com/openshiftmm/listinfo/users _______________________________________________ users mailing list [email protected] http://lists.openshift.redhat.com/openshiftmm/listinfo/users
