hussein-awala commented on PR #40568:
URL: https://github.com/apache/airflow/pull/40568#issuecomment-2209627933
> "why load the whole pod when you just need the status" is, i think, the
idea.
Unfortunately, in K8S, the pod status is the pod manifest, so the result of
`read_namespaced_pod_status` is the same as `read_namespaced_pod` with the same
performance/cost:
```python
import time
from kubernetes import client as k8s, config
config.load_kube_config(context="minikube")
client = k8s.ApiClient()
api = k8s.CoreV1Api(client)
namespace = "some namepsace"
pod_name = "some pod"
status = api.read_namespaced_pod_status(name=pod_name, namespace=namespace)
pod = api.read_namespaced_pod(name=pod_name, namespace=namespace)
print(pod == status)
```
the output of this script is:
```python
True
```
> but i'm curious -- can you tell us more what you're talking about with the
caching?
We have a caching mechanism in `read_pod` method:
```python
def read_pod(self):
_now = utcnow()
if (
self.read_pod_cache is None
or self.last_read_pod_at +
timedelta(seconds=self.read_pod_cache_timeout) < _now
):
self.read_pod_cache = self.pod_manager.read_pod(self.pod)
self.last_read_pod_at = _now
return self.read_pod_cache
```
I didn't dive into the code to check if we use/need it, but between two
methods that return the same result, and one of them has a cache mechanism, I
prefer the one with caching.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
