Python: Executing commands on a pod fails when only command kwarg is provided

[flaviamissi]

When calling connect_get_namespaced_pod_exec and providing only the required arguments + command, e.g. connect_get_namespaced_pod_exec("mypod", "namespace", command="ls"), I get the following exception:

ApiException: (0)
# Reason: Handshake status 404

After debugging it, I found that the issue lies on how the query string is built for multiple values vs single ones:
The resulting url path + query string for the request above is: /api/v1/namespaces/default/pods/mypod/exec&command=ls
Note the lack of ? when starting the query string.
But when giving another kwarg, e.g. connect_get_namespaced_pod_exec("mypod", "namespace", command="ls", stderr=True), the call works fine.

Note that this issue is specifically when using websockets. Relevant code is on websocket_call

Here's a full example that reproduces my problem:

from kubernetes.client import configuration
from kubernetes import config
from kubernetes.client.apis import core_v1_api

configuration.assert_hostname = False
config.load_kube_config()
c = core_v1_api.CoreV1Api()
c.connect_get_namespaced_pod_exec("mypod", "namespace", command="ls")
# raises exception "Handshake status 404"

FTR, here's where I first asked for/got help https://github.com/kubernetes-incubator/client-python/issues/36

[mildebrandt]

Using 3.0.0 from pip and python3.6, I get the same issue. However, the workaround of supplying another parameter does not work for me. Is there something else I could try to get this working? My resulting URL is a bit more complex, but even "ls" fails for me:

/api/v1/namespaces/demo/pods/mongo-0/exec?stderr=True&command=mongo&&command=admin&&command=--eval&&command=db.createUser%28%5C%7B+user%3A+%27myuser%27%2C+pwd%3A+%27mypass%27%2C+roles%3A+%5B%7B%27role%27%3A+%27userAdminAnyDatabase%27%2C+%27db%27%3A+%27admin%27%7D%5D+%5C%7D%29&

[flaviamissi]

Have you tried with 2.0.0? your url seems broken in a different way than mine was, you've got duplicated &, and my problem was that I got no ? at the beginning of the query string, so it makes sense that the work around doesn't fix it for you:(

[mildebrandt]

It seems my issue was the lack of the "container" parameter since there are two in my pod. However, I needed to run the command through curl to get that message back from the system:

HTTP/1.1 400 Bad Request
Content-Type: application/json
Date: Wed, 30 Aug 2017 02:20:52 GMT
Content-Length: 231

{
  "kind": "Status",
  "apiVersion": "v1",
  "metadata": {},
  "status": "Failure",
  "message": "a container name must be specified for pod mongo-0, choose one of: [mongo mongo-sidecar]",
  "reason": "BadRequest",
  "code": 400
}

It doesn't seem like the double ampersands cause any issue, but it's easily fixed by changing:

url += "&command=%s&" % quote_plus(command)

to

url += "&command=%s" % quote_plus(command)

in ws_client.py.

[mbohlool]

@eyeofthefrog nice catch. can you send a PR to fix ws_client.py?

[mildebrandt]

Yeah, I'll create one after my vacation.

[mildebrandt]

Done https://github.com/kubernetes-incubator/client-python/pull/346

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[fejta-bot]

Stale issues rot after 30d of inactivity.
Mark the issue as fresh with /remove-lifecycle rotten.
Rotten issues close after an additional 30d of inactivity.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle rotten

[fejta-bot]

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

[k8s-ci-robot]

@fejta-bot: Closing this issue.

In response to this:

Rotten issues close after 30d of inactivity.
Reopen the issue with /reopen.
Mark the issue as fresh with /remove-lifecycle rotten.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.