Add the node name in logs

In cluster mode (DCE), there's no indication whether the logs are coming from a node or another.
We should add the node name in the logs.

Options:

1. Add the name of the node as part of the filename - it would make the file name specific to the DCE and can be confusing in term of documentation and support
2. Add it as part of the taskID/requestID field in the logs - the current log format would remain compatible but it would mix different information in a same "field"
   Ex:

   2018.07.18 17:35:14 INFO  ce[vmh1-5CVSAaCpibHQ|AWSiOrRM0wkYktcHzt00][o.s.c.t.CeWorkerImpl] Execute task | project=multi-language-multi-module-project | type=REPORT | id=AWSiOrRM0wkYktcHzt00
   

   where vmh1-5CVSAaCpibHQ is the node name.
   Note that the "/" seems to be a valid character in requestID and can't be used.

3. Add it as a new dedicated field in the logs - it would break the current log format but, at the end, would make it clearer in case a tool is used to aggregate logs
   Ex:

   2018.07.18 17:35:14 INFO vmh1-5CVSAaCpibHQ ce[AWSiOrRM0wkYktcHzt00][o.s.c.t.CeWorkerImpl] Execute task | project=multi-language-multi-module-project | type=REPORT | id=AWSiOrRM0wkYktcHzt00
   

   or

   2018.07.18 17:35:14 INFO ce[vmh1-5CVSAaCpibHQ][AWSiOrRM0wkYktcHzt00][o.s.c.t.CeWorkerImpl] Execute task | project=multi-language-multi-module-project | type=REPORT | id=AWSiOrRM0wkYktcHzt00
   

The default naming of nodes (ex: sonarqube-15f1d6be-d9bf-4bd6-94e7-53c7ce19f3cd) should be adapted for something shorter. Node names need to be unique only in the context of a customer.