hack/scripts-dev: share docker image between test cases, clean up DNS tests #8958
Conversation
Signed-off-by: Gyu-Ho Lee <gyuhox@gmail.com>
…s-wildcard-run Signed-off-by: Gyu-Ho Lee <gyuhox@gmail.com>
| cat /dev/null >/etc/hosts | ||
|
|
||
| goreman -f /certs-wildcard/Procfile start & | ||
| sleep 7s |
There was a problem hiding this comment.
sleep in code always seems hacky.
There was a problem hiding this comment.
You need sleep until cluster is ready in this case.
There was a problem hiding this comment.
hmm, it seems like this test can be flaky since we don't really know when etcd cluster will be ready. if etcd cluster is rdy at 30s, the test will fail right?
There was a problem hiding this comment.
We don't run these in CIs anyway. For now, I just manually tested it and confirm that it works. Later when we integrate with CI, we can add some timeouts.
There was a problem hiding this comment.
@gyuho maybe add a TODO here to create some sort of waiting logic with timeout. if etcd cluster is ready, then go ahead. if not and has timeout, exit with error.
| --endpoints=https://m1.etcd.local:2379 \ | ||
| endpoint health --cluster | ||
|
|
||
| sleep 2s |
| cat /dev/null >/etc/hosts | ||
|
|
||
| goreman -f /certs/Procfile start & | ||
| sleep 7s |
|
any ideas on why jenkins are failing? |
Seems like jenkins plugin issue. |
|
@gyuho yeah, the code lgtm. we can add the timeout logic when we integrate into ci. |
Signed-off-by: Gyu-Ho Lee <gyuhox@gmail.com>
Clean up existing certs so that they can share the same docker image.
And address #8502.
Separate out from #8954.
/cc @fanminshi