Docker Community Forums

Share and learn in the Docker community.

UCP v0.6.0 container health red/green sign minor issue

(Jojojojo1234) #1

There is minor issue with UCP v0.6.0 in “containers” tab. Scenario is as follows:

stop && start existing container from command line (docker stop && docker start) from swarm node.
UCP will show container as “red” eventhough it is running - you can verify that either by docker ps -a or in UCP just drill into “Actions” section of that particular container and you see “Status running”

On the other hand - if you stop && start container from UCP it behaves correctly. However if you use “restart” the whole UCP page times out. Tested in Firefox 43.0 and Chrome 47.0.2526.106

I do not see any of these issues in UCP v0.5.0

(Vivek Saraswat) #2

Thanks Jojo, I also noticed the restart time-out bug as well and communicated it to the dev team. Should be fixed in next beta version. I’ll check in on the red/green sync issue as well.

(Thomas Barlow) #3

Hi jojojojo1234,

Would you be able to run a docker inspect <container-id> | grep Status against the Swarm master or UCP Controller? I’d like to see what the value of Status is that we’re receiving back from Swarm’s API.

Need a little more information here if you don’t mind, as I’m trying to reproduce this issue. To clarify, you have a multi-node Swarm cluster running. You’re then manually logging into a node and restarting a container via the local unix socket?

(Jojojojo1234) #4

hi Thomas,

Update on this: so, this issue is not happing as described above, my bad.Sorry. But this issue is there after I restarted docker engine on all nodes of swarm cluster (because i needed docker engine to start with overlay network startup options). So after you restart all nodes of swarm cluster you will find many system containers (like ucp-controller, ucp-kv, ucp-swarm-join, etc) marked red altought they are running. See this:

… all those marked red are actually running:

root@ThinkPad:/# for i in 1f9e81a23979 cfb56d012ad6 bba290f16654 9938787833b8 ccc03eb167d8 5ecca1da797b ; do echo $i; docker inspect $i | grep Status; done
"Status": “running”,
"Status": “running”,
"Status": “running”,
"Status": “running”,
"Status": “running”,
"Status": “running”,

… refreshing the page or logging out/in does not help. Also using different browser does not help

FYI: this is 6 node UCP/Swarm cluster: 1x master, 2x master replicas, 3 minions