[BUG] error starting session on aws k8

**Describe the bug**
```
import graphscope
from graphscope.framework.loader import Loader
from graphscope.dataset.ogbn_mag import load_ogbn_mag

# Set up the GraphScope session
graphscope_session = graphscope.session(cluster_type='k8s', show_log=True)

# Load the dataset
g = load_ogbn_mag(graphscope_session)

# Process the graph
# Here we create a simple PageRank application as an example
pr = graphscope.pagerank(g, delta=0.85, max_round=10)
pr = pr.unfold()
```

trying to launch a session using above code but am getting

```
ConnectionError: Connect coordinator timeout, code: UNAVAILABLE, details: failed
to connect to all addresses; last error: UNKNOWN: ipv4:192.168.3.241:30243:
Failed to connect to remote host: Operation timed out
```

**Expected behavior**
launch and connect to session

**Environment (please complete the following information):**
- GraphScope version: 0.23.0
- OS: Linux
- Kubernetes Version 1.24, EKS

Start by reproducing the graphscope.session(cluster_type='k8s') entry point with the provided GraphScope 0.23.0 example on Kubernetes 1.24/EKS. Investigate the coordinator connection timeout at 192.168.3.241:30243; done means the session launches and connects successfully.

Written by the indexing model from the issue text.