spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Fernando O." <fot...@gmail.com>
Subject Newbie Question
Date Thu, 11 Dec 2014 15:50:06 GMT
Hi guys,
    I'm planning to use spark on a project and I'm facing a problem, I
couldn't find a log that explains what's wrong with what I'm doing.

I have 2 vms that run a small hadoop (2.6.0) cluster. I added a file that
has a 50 lines of json data

Compiled spark, all tests passed, I run some simple scripts and then a yarn
example from the page:

./bin/spark-submit --class org.apache.spark.examples.SparkPi \
    --master yarn-cluster \
    --num-executors 3 \
    --driver-memory 512m \
    --executor-memory 512m \
    --executor-cores 1 \
    lib/spark-examples*.jar \
    10

they all worked, so now It was time to start playing with yarn integration

I did
./bin/spark-shell  -Dspark-cores-max=1 -Dspark.executor.memory=128m
 --driver-memory 385m --executor-memory 385m --master yarn-client


then

*scala> val file = sc.textFile("hdfs:///logs/events.log")*
14/12/11 09:50:35 INFO storage.MemoryStore: ensureFreeSpace(82156) called
with curMem=0, maxMem=278302556
14/12/11 09:50:35 INFO storage.MemoryStore: Block broadcast_0 stored as
values in memory (estimated size 80.2 KB, free 265.3 MB)
14/12/11 09:50:35 INFO storage.MemoryStore: ensureFreeSpace(17472) called
with curMem=82156, maxMem=278302556
14/12/11 09:50:35 INFO storage.MemoryStore: Block broadcast_0_piece0 stored
as bytes in memory (estimated size 17.1 KB, free 265.3 MB)
14/12/11 09:50:35 INFO storage.BlockManagerInfo: Added broadcast_0_piece0
in memory on 172.17.42.1:53158 (size: 17.1 KB, free: 265.4 MB)
14/12/11 09:50:35 INFO storage.BlockManagerMaster: Updated info of block
broadcast_0_piece0
file: org.apache.spark.rdd.RDD[String] = hdfs:///logs/events.log
MappedRDD[1] at textFile at <console>:12

*scala> val newsletters = file.filter(line =>
line.contains("newsletter_id"))*
newsletters: org.apache.spark.rdd.RDD[String] = FilteredRDD[2] at filter at
<console>:14

*scala> newsletters.count()*
14/12/11 09:50:47 INFO mapred.FileInputFormat: Total input paths to process
: 1
14/12/11 09:50:47 INFO spark.SparkContext: Starting job: count at
<console>:17
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Got job 0 (count at
<console>:17) with 2 output partitions (allowLocal=false)
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Final stage: Stage 0(count
at <console>:17)
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Parents of final stage:
List()
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Missing parents: List()
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Submitting Stage 0
(FilteredRDD[2] at filter at <console>:14), which has no missing parents
14/12/11 09:50:47 INFO storage.MemoryStore: ensureFreeSpace(2688) called
with curMem=99628, maxMem=278302556
14/12/11 09:50:47 INFO storage.MemoryStore: Block broadcast_1 stored as
values in memory (estimated size 2.6 KB, free 265.3 MB)
14/12/11 09:50:47 INFO storage.MemoryStore: ensureFreeSpace(1687) called
with curMem=102316, maxMem=278302556
14/12/11 09:50:47 INFO storage.MemoryStore: Block broadcast_1_piece0 stored
as bytes in memory (estimated size 1687.0 B, free 265.3 MB)
14/12/11 09:50:47 INFO storage.BlockManagerInfo: Added broadcast_1_piece0
in memory on 172.17.42.1:53158 (size: 1687.0 B, free: 265.4 MB)
14/12/11 09:50:47 INFO storage.BlockManagerMaster: Updated info of block
broadcast_1_piece0
14/12/11 09:50:47 INFO scheduler.DAGScheduler: Submitting 2 missing tasks
from Stage 0 (FilteredRDD[2] at filter at <console>:14)
14/12/11 09:50:47 INFO cluster.YarnClientClusterScheduler: Adding task set
0.0 with 2 tasks
14/12/11 09:50:47 INFO util.RackResolver: Resolved hdfs1 to /default-rack
14/12/11 09:51:02 WARN cluster.YarnClientClusterScheduler: Initial job has
not accepted any resources; check your cluster UI to ensure that workers
are registered and have sufficient memory
14/12/11 09:51:17 WARN cluster.YarnClientClusterScheduler: Initial job has
not accepted any resources; check your cluster UI to ensure that workers
are registered and have sufficient memory


On the node executing

The spark console shows the task:


Active Stages (1)Stage IdDescriptionSubmittedDurationTasks: Succeeded/Total
InputShuffle ReadShuffle Write0(kill
<http://localhost:4040/stages/stage/kill?id=0&terminate=true>)count at
<console>:17 <http://localhost:4040/stages/stage?id=0&attempt=0>+details

2014/12/11 10:18:255.8 min
0/2


In the Hadoop node executing the task:
NodeManager informationTotal Vmem allocated for Containers3.10 GBVmem
enforcement enabledtrueTotal Pmem allocated for Container1.48 GBPmem
enforcement enabledtrueTotal VCores allocated for Containers1
NodeHealthyStatustrueLastNodeHealthTimeThu Dec 11 10:17:49 ART 2014
NodeHealthReportNode Manager Version:2.6.0 from
e3496499ecb8d220fba99dc5ed4c99c8f9e33bb1 by jenkins source checksum
7e1415f8c555842b6118a192d86f5e8 on 2014-11-13T21:17ZHadoop Version:2.6.0
from e3496499ecb8d220fba99dc5ed4c99c8f9e33bb1 by jenkins source checksum
18e43357c8f927c0695f1e9522859d6a on 2014-11-13T21:10Z




Cluster MetricsApps SubmittedApps PendingApps RunningApps CompletedContainers
RunningMemory UsedMemory TotalMemory ReservedVCores UsedVCores TotalVCores
ReservedActive NodesDecommissioned NodesLost NodesUnhealthy NodesRebooted
Nodes101011 GB2.95 GB0 B1202 <http://hdfs0:8088/cluster/nodes>0
<http://hdfs0:8088/cluster/nodes/decommissioned>0
<http://hdfs0:8088/cluster/nodes/lost>0
<http://hdfs0:8088/cluster/nodes/unhealthy>0
<http://hdfs0:8088/cluster/nodes/rebooted>
Show 20406080100 entries
Search:
ID
User
Name
Application Type
Queue
StartTime
FinishTime
State
FinalStatus
Progress
Tracking UI
application_1418303626573_0001
<http://hdfs0:8088/cluster/app/application_1418303626573_0001>foteroSpark
shellSPARKdefaultThu, 11 Dec 2014 13:18:02 GMTN/ARUNNINGUNDEFINED
ApplicationMaster <http://hdfs0:8088/proxy/application_1418303626573_0001/>



I got the same results (but seeing the  Initial job has not accepted an...
on hadoop logs)

import org.apache.spark.{SparkConf, SparkContext}

/** This is a simple test*/
object TestSpark {
  def main (args: Array[String]) {
    System.err.println("Let's start")
    val sparkConf = new SparkConf().setAppName("TestSpark")
    val sc = new SparkContext(sparkConf)
    val file = sc.textFile("hdfs:///logs/events.log")
    val newsletters = file.filter(line => line.contains("newsletter_id"))
    val count = newsletters.count()
    System.err.println("count: " + count)
  }
}

*hadoop-2.6.0/logs/userlogs/application_1418303626573_0007/container_1418303626573_0007_01_000002*
> cat stderr
...
14/12/11 12:32:09 INFO storage.MemoryStore: ensureFreeSpace(2664) called
with curMem=236790, maxMem=280248975
14/12/11 12:32:09 INFO storage.MemoryStore: Block broadcast_1 stored as
values in memory (estimated size 2.6 KB, free 267.0 MB)
14/12/11 12:32:09 INFO storage.MemoryStore: ensureFreeSpace(1682) called
with curMem=239454, maxMem=280248975
14/12/11 12:32:09 INFO storage.MemoryStore: Block broadcast_1_piece0 stored
as bytes in memory (estimated size 1682.0 B, free 267.0 MB)
14/12/11 12:32:09 INFO storage.BlockManagerInfo: Added broadcast_1_piece0
in memory on hdfs0:51045 (size: 1682.0 B, free: 267.2 MB)
14/12/11 12:32:09 INFO storage.BlockManagerMaster: Updated info of block
broadcast_1_piece0
14/12/11 12:32:09 INFO scheduler.DAGScheduler: Submitting 2 missing tasks
from Stage 0 (FilteredRDD[2] at filter at TestSpark.scala:10)
14/12/11 12:32:09 INFO cluster.YarnClusterScheduler: Adding task set 0.0
with 2 tasks
14/12/11 12:32:09 INFO util.RackResolver: Resolved hdfs1 to /default-rack
14/12/11 12:32:24 WARN cluster.YarnClusterScheduler: Initial job has not
accepted any resources; check your cluster UI to ensure that workers are
registered and have sufficient memory
14/12/11 12:32:39 WARN cluster.YarnClusterScheduler: Initial job has not
accepted any resources; check your cluster UI to ensure that workers are
registered and have sufficient memory
14/12/11 12:32:54 WARN cluster.YarnClusterScheduler: Initial job has not
accepted any resources; check your cluster UI to ensure that workers are
registered and have sufficient memory



Can anyone give me a tip of what I'm missing? :D


Thanks!

Mime
View raw message