failed to start container when scanbox isn't running
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
Zun |
Triaged
|
Low
|
Unassigned |
Bug Description
1,
[root@tfg-104 zun(keystone_
+------
| uuid | name | image | status | task_state | addresses | ports |
+------
| 838f35f2-
+------
2,
[root@tfg-104 zun(keystone_
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
c6820a4093d1 test:1 "supervisord -c /etc/" 18 hours ago Exited (128) 24 minutes ago zun-838f35f2-
c3ec9b47d833 kubernetes/pause "/pause" 18 hours ago running zun-sandbox-
3,
docker stop c3ec9b47d833
4,
[root@tfg-104 zun(keystone_
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
c6820a4093d1 test:1 "supervisord -c /etc/" 18 hours ago Exited (128) 24 minutes ago zun-838f35f2-
c3ec9b47d833 kubernetes/pause "/pause" 18 hours ago Exited (128) 24 minutes ago zun-sandbox-
5,
zun start test
6,
[root@tfg-104 zun(keystone_
+------
| uuid | name | image | status | task_state | addresses | ports |
+------
| 838f35f2-
+------
After starting the container, the status is still Stopped, but there's no error.
I think we should check the status of scanbox, if it's not running, we should raise a error.
Changed in zun: | |
assignee: | nobody → feng.shengqin@zte.com.cn (feng-shengqin) |
@Feng,
This is an interesting bug report. I agree that we need to handle this counter case in "start" . Perhaps, the "start" command should check the status of the sandbox first and start it if it is "stopped"? Eventually, I agree with you that Zun should return an error if we are not able to start the sandbox.