Log Pumba "kill" activities with more details about affected containers

[alexei-led]

Boris proposal (Slack):

The decision about if the test passed or failed may be not easy and I think Pumba can help with this by logging its activities (08/03 12:30 Pumba killed container "abc") and also docker's activities in general (08:03 12:32 container "xyz" has stopped) so that user can build the events list in timed manner and correlate container level activities with application level logs to understand what really happened in the system and how long it took each phase of the process. It may be covered by the debug option but it is not clear from the description. It may also send messages like this to Slack or any other destination using webhooks. This is especially useful since Pumba looking only on current host dockers, having this kind of logging reported to common storage gives a cluster-level picture when impacted service could be running remotely.

[alexei-led]

More detailed log is available. It's possible also to log in JSON Logstash/Splunk friendly format. It's also possible to send events to Slack channel (need to check this too)