Reputation: 65
We are working on a stack that involves running google chrome extensions inside docker containers. The setup works flawless for less number of containers(around 10-15). However as this number is increased I randomly see a few containers where chrome has failed to start.
After getting hold of the log files this is what I managed to extract. I did a compare from a sample of a working and non-working example and these are the results. The left section is the non-working log.
https://www.diffnow.com/?report=ilkum
Can someone point out what exactly is going wrong?
Upvotes: 0
Views: 865
Reputation: 61571
Something is happening at the lxc/docker level. Did you check your /var/log/docker.log file while this is happening?
So, if you take a look at this
You see that the additional message you are seeing:
[8:47:0825/154356:VERBOSE1:io_thread.cc(397)] Observed a change to the network IP addresses
[8:47:0825/154356:VERBOSE1:io_thread.cc(408)] Observed a change to network connectivity state CONNECTION_ETHERNET
is because the ethernet interface is flapping. So it could be that your container is flapping it's eth interface as it's coming up because of an error. (Maybe, running out of memory)
Upvotes: 1