Checkpointing of stateful operators in spark streaming -

i developing streaming application uses mapwithstate function internally...

i need set checkpointing interval of checkpoinitnd data manually..

this sample code..

var newcontextcreated = false      // flag detect whether new context  created or not  // function create new streamingcontext , set def creatingfunc(): streamingcontext = {    // create streamingcontext val ssc = new streamingcontext(sc, seconds(batchintervalseconds))    // create stream generates 1000 lines per second   val stream = ssc.receiverstream(new dummysource(eventspersecond))    // split lines words, , create paired (key-value) dstream   val wordstream = stream.flatmap { _.split(" ")  }.map(word => (word, 1))     // represents emitted stream trackstatefunc. since    emit every input record updated value,   // stream contain same # of records input dstream.   val wordcountstatestream = wordstream.mapwithstate(statespec)  wordcountstatestream.print()    // snapshot of state current batch. dstream contains 1 entry per key.   val statesnapshotstream = wordcountstatestream.statesnapshots()    statesnapshotstream.foreachrdd { rdd => rdd.todf("word", "count").registertemptable("batch_word_count") }     ssc.remember(minutes(1))  // make sure data not deleted time     query interactively   ssc.checkpoint("dbfs:/streaming/trackstate/100")   println("creating function called create new streamingcontext")      newcontextcreated = true        ssc     }