Kubernetes Network Attack failing using DNS names

Created by Jamie Martin, Modified on Tue, 18 Oct 2022 at 04:49 PM by Jamie Martin

Use case: If you are attempting to run a network attack, from one service to another, using the hostname field to specify a particular service to apply the impact to.


For example, if you want to apply 100ms latency on all traffic from Service123 to ServiceABC. You are able to supply the DNS name of the service in our hostname field to apply the latency. You can use the following format: servicename.namespace if the services live in different namespaces. You can also use the FQDN, such as: cartservice.default.svc.cluster.local.


If you are seeing the following failure, in the execution logs, while attempting to use the DNS name: network: lookup failed for 'cartservice.default.svc.cluster.local': "failed to lookup address information: Name does not resolve". This can happen for hostnames that have no DNS records (NXDOMAIN), or for hostnames whose DNS records contain no IP Addresses, like NS records. 


 Please check your helm chart values with: helm get values gremlin -n gremlin


If there is a value of hostNetwork: true, please re-deploy the helm chart with hostNetwork = false (default value for our helm chart). Re-run the attack and you should now see a successful execution.


 


For more information, on helm chart values, please see our Helm GitHub repository here: https://github.com/gremlin/helm/tree/master/gremlin

Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select atleast one of the reasons

Feedback sent

We appreciate your effort and will try to fix the article