Load Charmader Utilities



In [1]:

    
import org.att.charmander.CharmanderUtils

See what tasks are currently running and pick first test



In [8]:

    
val runningSimulators = CharmanderUtils.getMeteredTaskNamesFromRedis
val simulatorToInvestigate = runningSimulators(0)
runningSimulators









    Out[8]:





List(lookbusy80)

Retrieve memory-usage datapoints for the first simulator



In [3]:

    
val memoryUsage= CharmanderUtils.getRDDForTask(sc, simulatorToInvestigate, "memory_usage", 100)

Just for curiosity lets look at the first data point



In [4]:

    
memoryUsage.first









    Out[4]:





List(1425192395000, 4050001, 84971520)

Translate Datapoints in to a Spark-SQL Database



In [5]:

    
val sqlContext = new org.apache.spark.sql.SQLContext(sc)
import sqlContext._

case class MemoryUsage(timestamp: BigDecimal, memory: BigDecimal)

val memoryusageRdd = memoryUsage.map(p => MemoryUsage(BigDecimal(p(0).asInstanceOf[BigInt]), BigDecimal(p(2).asInstanceOf[BigInt])))
memoryusageRdd.registerTempTable("memoryusage")

Get max-memory usage



In [9]:

    
val newestMax = sqlContext.sql("select max(memory) from memoryusage").first()
newestMax









    Out[9]:





[84971520]

Set the new max value in our task intelligence database



In [7]:

    
CharmanderUtils.setTaskIntelligence(simulatorToInvestigate, "mem", newestMax(0).toString)

To Verify that the value is set

Visit redis and look for the task-intelligence section



In [20]: