Created 06-27-2016 07:34 PM
Created 06-27-2016 07:36 PM
Created 06-28-2016 06:57 PM
Changing AMS to distributed mode only means that AMS HBase will run separate daemons for Master and RegionServer. Both these processes will still be co-located with AMS collector.
Whereas, in embedded mode, AMS uses HBase in "standalone" mode - 1 daemon for HBase Master, RS and its own local Zk
Created 06-28-2016 07:19 PM
Well distributed mode means it should use cluster resources and not local resources rt? So Customer is asking whats the use of distributed mode and still have processes running locally?
Created 06-28-2016 07:55 PM
AMS in distributed mode does use the cluster resources. In distributed mode, AMS HBase is configured to write to HDFS on the cluster. In fact, it can be configured to write to any external file system as well. The AMS HBase processes though, run locally on the collector host.
Created 06-30-2016 08:02 PM
At this moment, 'embedded'/'distributed' decides only about where the AMS collected data is stored. Distributed only means the data collected by AMS is stored in HDFS and nothing else. It uses only the local resource except for HDFS. In embedded mode, the AMS data is stored in a physical file in the location defined in /etc/ams-hbase/conf/hbase-site.xml and resides local to the node where the collector is running.
Currently, there is only one Region Server in the Standalone Hbase used by AMS. Slowly AMS has started using cluster resources - for example, in the recent releases, the Cluster zookeeper could be used. At some point, we might have multiple region servers spread across the nodes in cluster as well.
Created 06-05-2017 07:05 AM
Created 06-05-2017 08:53 PM
Created 06-05-2017 10:03 PM