Here is a strange problem on our docker testing machines. We have one docker prod container that uses chat and workers and is behind a proxy. So port 8072 is accessible and longpollig proxied to it.
Now, there is a second docker container for testing that basically uses the same setup. It also has chat, workers, and longpolling configured and bound to host. Since that container is not proxied, I connect via ssh to docker host and forward the remote localhost ports to my laptop. These are actually different ports from the prod docker but for clarity I use the standard ones here. The problem also shows on the proxied container when circumventing the proxy by ssh fowarding localhost ports.
If I open localhost:8069 in my browser, I get the db selection. I can log in to any database and do stuff. But I cannot restore databases nor can I get good dumps. The restore gives me error 400 bad request. The dump produces incomplete zip files that are much smaller than they should be and do not contain all the data.
If I disable workers and connect to localhost:8069, I can dump restore just fine.
What puzzles me is this: when workers are enabled, I can point my browser to localhost:8072 and get the db selection view and then I can restore/dump datbases just fine.
Is there an explanation for this behaviour? I thought 8072 was completely optional and is not required for normal use. But here it looks like, given workers!=0, 8072 is the usable port while 8069 has massive problems.
Please try to give a substantial answer. If you wanted to comment on the question or answer, just use the commenting tool. Please remember that you can always revise your answers - no need to answer the same question twice. Also, please don't forget to vote - it really helps to select the best questions and answers!
About This Community
|Asked: 8/23/16, 9:33 AM|
|Seen: 227 times|
|Last updated: 9/2/16, 11:33 AM|