hi all   we have HDP version 2.6.5 and ambari 2.6.2.2   we configured the following    Metrics Service operation mode=distributed hbase.rootdir=hdfs://hdfsha:8020/user/ams/hbase   and all the rest according to - https://cwiki.apache.org/confluence/display/AMBARI/AMS+-+distributed+mode   so ambari configured to save the data on hdfs instead the local disk   after metrics collector service restart we saw the following from log     2020-02-20 20:09:15,770 INFO [timeline] timeline.HadoopTimelineMetricsSink: No live collector to send metrics to. Metrics to be sent will be discarded. This message will be skipped for the next 20 times. 2020-02-20 20:11:35,769 INFO [timeline] timeline.HadoopTimelineMetricsSink: No live collector to send metrics to. Metrics to be sent will be discarded. This message will be skipped for the next 20 times.     and     2020-02-20 20:30:10,306 INFO org.apache.hadoop.metrics2.impl.MetricsSystemImpl: phoenix metrics system started 2020-02-20 20:30:10,450 WARN org.apache.hadoop.hbase.io.util.HeapMemorySizeUtil: hbase.regionserver.global.memstore.upperLimit is deprecated by hbase.regionserver.global.memstore.size 2020-02-20 20:30:15,327 WARN org.apache.hadoop.hbase.io.util.HeapMemorySizeUtil: hbase.regionserver.global.memstore.upperLimit is deprecated by hbase.regionserver.global.memstore.size     and without metrics as       and after some time the metrics collector service failed    why this happens ?      NOTE - I must to say that - when data of metrics saved on local disk then everything was ok    is that because reaching data on HDFS is more slow? ... View more

hi all   we have HDP version - 2.6.4    with active and standby namenode   is it possible to capture the active namenode IP by REST API?     we try by the following API , but API not return any output    curl -sH "X-Requested-By: ambari" -u admin:admin -H "X-Requested-By: ambari" -X GET http://master:8080/api/v1/clusters//services/HDFS     ... View more

we are running spark streaming app that consumes data from kafka topics we want to know how to tune the parameter - `yarn.scheduler.capacity.maximum-am-resource-percent`   according to documentation:   yarn.scheduler.capacity.maximum-am-resource-percent: Maximum percent of resources in the cluster which can be used to run application masters i.e. controls number of concurrent running,   on some document we even see that recomneded to utilise it to `90 percent` for best results, but the default is `10%`   so my question is do we need to tune this parameter according to cluster size?   or what is the best practice in order to get good results? ... View more

hi all   how to verify if specific service required restart?   for example - HDFS service   in case HDFS service required restart then ambari will show this as     is it possible by rest api to understand if service required restart or not?    for example for - YARN / HIVE , etc ... View more