Does aggregation on metrics after grouping by these three attributes:
processorId
2. time section(represented as a index integer)
3. metricName(like sendThroughput)
It assumes that for each org.apache.gearpump.cluster.MasterToClient.HistoryMetricsItem, the
name follow the format app(appId).processor(processorId).task(taskId).(metricName)
It parses the name to get processorId and metricName. If the parsing fails, then current
org.apache.gearpump.cluster.MasterToClient.HistoryMetricsItem will be skipped.
Does aggregation on metrics after grouping by these three attributes:
It assumes that for each org.apache.gearpump.cluster.MasterToClient.HistoryMetricsItem, the name follow the format app(appId).processor(processorId).task(taskId).(metricName)
It parses the name to get processorId and metricName. If the parsing fails, then current org.apache.gearpump.cluster.MasterToClient.HistoryMetricsItem will be skipped.
NOTE: this class is optimized for performance.