Github user mmiklavc commented on the issue:
https://github.com/apache/metron/pull/1252
Pertaining to the topology args, our elasticsearch_master.py script always
passes in args based on the active topology. I also added a note for changing
the setting in Ambari as well as the manual process, per your suggestion, to
Upgrading.md.
```
# which enrichment topology needs started?
if self.__params.enrichment_topology == "Unified":
topology_flux =
"{0}/flux/enrichment/remote-unified.yaml".format(self.__params.metron_home)
topology_props =
"{0}/config/enrichment-unified.properties".format(self.__params.metron_home)
elif self.__params.enrichment_topology == "Split-Join":
topology_flux =
"{0}/flux/enrichment/remote-splitjoin.yaml".format(self.__params.metron_home)
topology_props =
"{0}/config/enrichment-splitjoin.properties".format(self.__params.metron_home)
else:
raise Fail("Unexpected enrichment topology; name=" +
self.__params.enrichment_topology)
# start the topology
start_cmd_template = """{0}/bin/start_enrichment_topology.sh
--remote {1} --filter {2}"""
Logger.info('Starting ' + self.__enrichment_topology)
start_cmd =
start_cmd_template.format(self.__params.metron_home, topology_flux,
topology_props)
Execute(start_cmd, user=self.__params.metron_user, tries=3,
try_sleep=5, logoutput=True)
```
---
