Replies: 1 comment
-
Hello again, I made a new test and add all steps here. Hope, somebody can confirm this? host 1 (manager):
host 2 (worker):
On host 1:
On host 2:
With some pc:
On host 1:
On host 2:
With some pc:
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hi,
I'm current running swarm in development with one manager + 3 workers. Manager (M) and 1 worker (W1) are on google cloud. 1 worker (W2) on hostingpalvelu virtual server and 1 on aws (W3). There is one overlay network where all services are connected to. All os are centos7. All behaving as same with docker 17.5 and 17.12 ce. I have previously this same problem, it solved, maybe update on some docker or centos package. Now I have it again.
I started one service on M. Container starts at M. Connect from browser works ok to M and W1 (both on google). From W2 and W3 I cannot connect. On those hosts I can run some small app, which is using same port and on that I can connect ok.
I drained M and it moves to W1. All works same as above.
I drained W1. service moved on W2. I cannot connect anymore from M or W1 and not still W3 ip+port. Same small app as above, cannot run anymore on W2 - port reserved. On W2 ip+port works ok.
I drained W2 and service moved to W3 and all works same as W2 above. Just connection from W3 p+port
Then I update M & W1 back to active and drained W3. Service moved to M and connection from M & W1 are ok, but W2 and W3 cannot been connected. Small app can run on W2 and W3, and it responsing on same port, which should be swarm?
At same time on workers, there is problem with docker dns ex. on W3, I cannot ping/connect service by name from container, which is running on M or W1 or W2.
Beta Was this translation helpful? Give feedback.
All reactions