Installed mapr on 3 nodes on aws instances running RHEL 6.5, nfs started on one node, no errors so far. Installed Kafka tested producer consumer in console, no errors here either. Installed Divolte, working well writing to kafka topics.
Now Adding strom to mapr cluster as per mapr documentation here Pretty clear and easy to follow, configured required settings in storm.yaml file (located at /opt/mapr/storm/storm-/conf).
I go to start the services with:
maprcli node services -name nimbus -action start -nodes mynode
maprcli node services -name storm-ui -action start -nodes mynode
maprcli node services -name supervisor -action start -nodes myothernode1 myothernode2 myothernode3
which returns nothing so I guess no news is good news. I am quite new to the hadoop ecosystem and I am working with a lot of partial knowledge.
My question is how do I add things like spark or storm to my mapr cluster. I followed the storm docs and storm does not appear in the mapr mcs on any of my nodes.
Tried: restarting mapr services, start the storm install from scratch (carefully) locate storm and nimbus show that its there under mapr roles but I don't understand what the next step is. Mapr docs say I should be able to see it in the mapr mcs. Do I need to run configure.sh? if so what parameters allow me to add mapr services?
Any help would be much appreciated
Aucun commentaire:
Enregistrer un commentaire