[CRIU] [Users] socket will take at least 0.5 seconds to recovery after docker restore done

[Pavel Emelyanov]

On 07/21/2015 06:25 PM, Yanbao Cui wrote:
> thanks Saied!
> The delay is happened after restore. The -existing- socket(s) cannot work, but at
> the same time ping the container IP is OK.
> 
> So I am confused that the delay is caused by Docker or CRIU? Because it seems that
> it only happened in Docker migration with CRIU

I would say that this is related to TCP checkpoint-restore. The restored connection
starts with zero window and sends the probe to remote side to get it's actual value.
This can be the reason for delay.

Can you collect tcpdump for your connection with timestamps starting from the moment
of restore so we could look at what packets are traveling?

-- Pavel