[systest] We should wait until Nailgun is become ready after resetting master node
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Fuel for OpenStack |
Fix Released
|
High
|
Artem Panchenko | ||
5.1.x |
Fix Released
|
High
|
Artem Panchenko | ||
6.0.x |
Fix Released
|
High
|
Fuel QA Team |
Bug Description
In system tests we destroy master node virtual machine and then start it if it's not accessible after snapshot revert:
https:/
But then test often fails with '502 error':
Because nailgun isn't ready yet (but nginx container is started already):
(venv-nailgun-
Domain env_base_
(venv-nailgun-
Domain env_base_
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
curl: (7) Failed to connect to 10.108.20.2 port 8000: Connection refused
(venv-nailgun-
<html>
<head><title>502 Bad Gateway<
<body bgcolor="white">
<center><h1>502 Bad Gateway<
<hr><center>
</body>
</html>
We should wait some time until Nailgun's API becomes available.
Fix proposed to branch: master /review. openstack. org/120985
Review: https:/