Question: as root?

Asked By
Asked At
2018-05-12 12:04:25

Found 15 possible answers.

User Answered At Possible Answer
hsaadat 2018-05-12 12:05:48 okay i just ran as sudo i will ask on the kops channel. thanks @embik I need to start the daemon manually so same error
troylevin 2018-05-12 13:18:14 @troylevin uploaded a file: Kube-DNS No Endpoints All - hoping someone can help explain why my Kube-DNS Service won't attach the pods as an endpoint
vaizki 2018-05-12 13:21:31 Endpoints: ok and now I saw your second comment... Doesn’t look like the DNS pods are being picked up by the service
troylevin 2018-05-12 13:26:14 @vaizki - the service seems to be working. options ndots:5 search kube-system.svc.cluster.local svc.cluster.local cluster.local nameserver [root@net-tools01-7999d5cc59-w5j5t /]# cat /etc/resolv.conf @vaizki ok [root@net-tools01-7999d5cc59-w5j5t /]# curl @vaizki I mean working in that I can hit it from another POD
msau42 2018-05-12 13:32:45 Awesome glad to hear it's working
blazhe.mitrev 2018-05-12 13:39:25 I have solved the problem with this link there was a lack of permissions
phin 2018-05-12 14:08:46 Is it possible to assign multiple external IPs per node in kubernetes engine?
upadhyay.ms01 2018-05-12 14:56:35 @phin egress router is
davelund 2018-05-12 14:58:57 is there a way to make a pods hostname resolve to rather than to its SDN ip?
troylevin 2018-05-12 15:07:14 @vaizki @chancez ALL. Any other ideas why my DNS service doesn't pick up any endpoint? I've confirmed that the selectors match the lables. I've confirmed that all the readiness probes are working for the 3 containers in the kube-dns POD. ok root@net-tools01-7999d5cc59-w5j5t /]# curl kube-dns-6f5f6f74-n2p8q 3/3 Running 0 21h k8s-app=kube-dns,pod-template-hash=29192930[r NAME READY STATUS RESTARTS AGE LABELS kubectl get pods kube-dns-6f5f6f74-n2p8q --namespace=kube-system --show-labels
marc.muller 2018-05-12 15:09:04 hey guys, I'm trying to get aquainted with kubectl. Any ideas why 'kubectl run debian-7 --image=debian:latest --port 2227 /bin/bash' results in CrashLoopBackoff while 'kubectl run -i debian-6 --image=debian:latest --port 2226 /bin/bash' seems to work just fine?
liggitt 2018-05-12 15:10:34 you dropped -i in the first invocation, right? so it's not interactive, so bash immediately exits
hubt 2018-05-12 15:13:28 you can overwrite the /etc/hosts file does kubectl describe pod show health check failures?
robinlinacre 2018-05-12 15:17:44 hi all. I have a set of data files which I need to process once a day. They can be processed in parallel. I’m looking for the easiest way of setting up a job where each pod is ‘aware’ of which file it should process. I have got this working using the method in this example: but running a queue seems very convoluted in my case, where I have a fixed array of datasets. The docs have a section on parallel jobs with a fixed completion count which has ‘not implemented yet: each pod passed a different index in the range 1 to .spec.completions.’ This sounds very much like what I need. Is there another similar option?
marc.muller 2018-05-12 15:19:12 liggitt: exactly... does kubectl consider a container running only as long as the terminal stays open?

Related Questions