Fixed issue where Connection Refused Log error printed incorrect port number

[hvaghani221]

Fixed splunk/splunk-connect-for-kubernetes#371 and #184

Exception message was changed in between stack trace and at some point it was replaced with proxy_port which defaults to 80. So raised cause of exception Net::HTTP::Persistent::Error is raised.

Now, the following error will be logged:

2021-12-09 08:59:16 +0000 [error]: #0 failed to flush the buffer, and hit limit for retries. dropping all chunks in the buffer queue. retry_times=5 records=29841 error_class=Errno::ECONNREFUSED error="Failed to open TCP connection to 10.202.9.155:12088 (Connection refused - connect(2) for \"xx.xx.xx.xx" port yyyy)"