Fix local_e2e_tests CI job

[beni0888]

Description of the change

CI is broken at local_e2e_tests in the kubeapps-general workflow for the following reasons:

• It's failing at the Export cluster variables step because it's unable to get the DEX IP. The error happens because docker network inspect kind| jq '.[0].IPAM.Config returns where the item we are looking for has been moved from the first to the second position ([0] -> [1]).
• It's failing for the flux test group because it's unable to install Flux due to a 404 response from the server.

Benefits

The CI works again.

Possible drawbacks

This is not the ideal solution because is fragile. If the order of the items in the array change again the future, the CI will be broken again.

Applicable issues

• fixes #

Additional information

Sample failing job: https://github.com/vmware-tanzu/kubeapps/actions/runs/9850222972/job/27223662707

DEFAULT_DEX_IP=172.18.0.2
DEX_IP=null...1
ADDITIONAL_CLUSTER_IP=null...2
Context "kind-kubeapps-ci" modified.
Default IP does not match with current IP used in Kind
Error: Process completed with exit code 1.

This is the output of docker network inspect kind| jq '.[0].IPAM':

runner@fv-az573-34:~/work/kubeapps/kubeapps$ docker network inspect kind| jq '.[0].IPAM'
{
  "Driver": "default",
  "Options": {},
  "Config": [
    {
      "Subnet": "fc00:f853:ccd:e793::/64"
    },
    {
      "Subnet": "172.18.0.0/16",
      "Gateway": "172.18.0.1"
    }
  ]
}
<!-- If there's anything else that's important and relevant to your pull
request, mention that information here.-->

[netlify]

✅ Deploy Preview for kubeapps-dev canceled.

Built without sensitive environment variables

Name	Link
🔨 Latest commit	daa44ef
🔍 Latest deploy log	https://app.netlify.com/sites/kubeapps-dev/deploys/6694c135e99aaa0008dee3dc