Description of problem:
 Changing value for reconnect-timeout doesn't have any effect unless we reload a server.

Version-Release number of selected component (if applicable):
 EAP 6.4.0.DR3

How reproducible:
 Always

Steps to Reproduce:
1. Configure a syslog handler
/core-service=management/access=audit/logger=audit-log:write-attribute(name=enabled,value=true)
batch
/core-service=management/access=audit/syslog-handler=mysyslog:add(formatter=json-formatter, max-failure-count=2)
/core-service=management/access=audit/syslog-handler=mysyslog/protocol=tcp:add(host=<syslog_host>, port=514, reconnect-timeout=30)
run-batch
/core-service=management/access=audit/logger=audit-log/handler=mysyslog:add

2. Change the value for reconnect-timeout
./core-service=management/access=audit/syslog-handler=mysyslog/protocol=tcp:write-attribute(name=reconnect-timeout, value=20)

3. Make a syslog server failover and see logs

Actual results:
 JBAS013417: Attempting reconnect to syslog handler 'mysyslog; after timeout of 30 seconds

Expected results:
 JBAS013417: Attempting reconnect to syslog handler 'mysyslog; after timeout of 20 seconds
or
 "reload required"

Additional info:
 First mgmt operation on disabled syslog server doesn't increase the failure counter, every other triggers NPE.