Contains different configurations and procedures to enable logging on different daemons on AWS EMR cluster.
[Please contribute to this article to add additional ways to enable logging]
HBASE on S3 :
1 2 3 4 5 6 7 | { "Classification": "hbase-log4j", "Properties": { "log4j.logger.com.amazon.ws.emr.hadoop.fs.shaded.com.amazonaws.request" : "DEBUG", "log4j.logger.com.amazon.ws.emr.hadoop.fs.shaded.com.amazonaws.latency" : "ERROR" } } |
This will enable calls made from EMRFS from HBASE.
Important to troubleshoot S3 consistency issues and failures for HBASE on S3 cluster.
Enabling DEBUG on Hive Metastore daemon (its Datastore) on EMR :
1 2 3 4 5 6 7 8 9 10 11 12 13 | vim /etc/hive/conf/hive-log4j2.properties
status = DEBUG name = HiveLog4j2
logger.DataNucleus.name = DataNucleus logger.DataNucleus.level = DEBUG
logger.Datastore.name = Datastore logger.Datastore.level = DEBUG
sudo stop hive-hcatalog-server - sudo start hive-hcatalog-server |
or
1 2 3 4 5 6 7 | { "Classification": "hive-log4j2", "Properties": { "logger.Datastore.level" : "DEBUG", "logger.DataNucleus.level" : "INFO" } } |
Logs at /var/log/hive/user/hive/hive.log
HUE:
use_get_log_api=true
in the beeswaxsection
of the hue.ini
configuration file.
Hadoop and MR :
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 | [ { "classification": "hadoop-log4j", "properties": { "log4j.logger.com.amazonaws": "DEBUG", "log4j.logger.com.amazonaws.http.AmazonHttpClient": "DEBUG", "log4j.logger.org.apache.hadoop.fs.s3a.S3AFileSystem": "DEBUG", "log4j.logger.emr": "DEBUG", "hadoop.root.logger": "DEBUG,console" } }, { "Classification": "mapred-env", "Configurations": [ { "Classification": "export", "Configurations": [], "Properties": { "HADOOP_MAPRED_ROOT_LOGGER": "DEBUG,DRFA" } } ], "Properties": {} } ] |
Enable GC verbose on Hive Server 2 JVM:
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 | #!/bin/bash
# This script sources the hive-env.sh on EMR 4.x.x/5.x.x with custom HADOOP_OPTS for hiveserver2 and restarts the HS2 process. # These options enable GC verbose on HS2 which gets logged to /var/log/hive/hive-server2.out. # On OOM , just before HS2 gets killed(with kill -9 command) , it also issues a heap dump(to /var/log/hive path).
echo '' | sudo tee --append /usr/lib/hive/conf/hive-env.sh
echo 'if [ "$SERVICE" = "hiveserver2" ] then export HADOOP_OPTS="$HADOOP_OPTS -server -verbose:gc -XX:+PrintGCDetails -XX:+PrintGCTimeStamps -XX:+PrintGCDateStamps -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=/var/log/hive" fi' | sudo tee --append /usr/lib/hive/conf/hive-env.sh
echo "stopping HS2" sudo stop hive-server2 sleep 5; echo "starting HS2" sudo start hive-server2 |
WIRE OR DEBUG logging on EMR to check calls to S3 and DDB for DynamoDb connector library :
Paste the following on log4j configurations of Hadoop / hive / spark etc.
/etc/hadoop/conf/log4j.properties
/etc/hadoop/conf/container-log4j.properties
/etc/hive/conf/hive-log4j2.properties
/etc/spark/conf/..
1 2 3 4 5 | log4j.rootCategory=DEBUG, stdout log4j.appender.stdout=org.apache.log4j.ConsoleAppender log4j.appender.stdout.layout=org.apache.log4j.PatternLayout log4j.appender.stdout.layout.ConversionPattern=%d{ABSOLUTE} %5p %t %c{2} - %m%n log4j.logger.org.apache.hadoop.hive=ALL |
Debug on S3 Calls from EMR HIVE :
These metrics can be obtained from the hive.log
when enabling debug logging in aws-java-sdk. To enable this logging, add the following line to '/etc/hive/conf/hive-log4j.properties'
. The Configuration API can be used as well.
1 | log4j.logger.com.amazonaws=DEBUG |
1 2 3 4 5 6 7 8 | To count the number of HEAD calls made execute:
$ cat /var/log/hive/tmp/hadoop/hive.log | grep "Sending Request: HEAD" | wc -l 4
To count the number of GET calls made execute:
$ cat /var/log/hive/tmp/hadoop/hive.log | grep "Sending Request: GET" | wc -l |
Enable DEBUG logging for Http Connection pool:
(from spark) by adding the following to /etc/spark/conf/log4j.properties
1 | log4j.logger.com.amazon.ws.emr.hadoop.fs.shaded.org.apache.http.impl.conn.PoolingHttpClientConnectionManager=DEBUG |
*Tez overwrites the loglevel options we have passed. Please see the related items.*
1 2 3 4 5 6 7 8 9 | { "Classification": "tez-site", "Properties": { "tez.task.log.level" : "DEBUG", "tez.am.log.level" : "DEBUG", "tez.root.logger" : "DEBUG,CLA", "tez.task-specific.launch.cmd-opts" : "-Dlog4j.configuration=log4j.properties" } } |
Enabling Debug on Hadoop log to log calls by EMRFS :
/etc/hadoop/conf/log4j.properties
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 | hadoop.root.logger=DEBUG,console
# Jets3t library
log4j.logger.org.jets3t.service.impl.rest.httpclient.RestS3Service=DEBUG
# AWS SDK & S3A FileSystem
log4j.logger.com.amazonaws.http.AmazonHttpClient=DEBUG
log4j.logger.org.apache.hadoop.fs.s3a.S3AFileSystem=WARN
log4j.logger.com.amazonaws=DEBUG
log4j.logger.com.amazon.ws.emr.hadoop.fs.shaded.com.amazonaws.request=DEBUG
log4j.logger.com.amazon.ws.emr.hadoop.fs.s3.lite.handler.RequestIdLogger=DEBUG #logs the AWS request id and S3 "extended request id" from each response |
You can use same logging config for other Application like spark/hbase using respective log4j config files as appropriate. You can also use EMR log4j configuration classification like hadoop-log4j or spark-log4j to set those config’s while starting EMR cluster.(see below for sample JSON for configuration API)
https://docs.aws.amazon.com/emr/latest/ReleaseGuide/emr-configure-apps.html
DEBUG on EMR Logpusher Logs :
Edit this file on Master / Slave’s manually and restart Logpusher.
/etc/logpusher/logpusher-log4j.properties
1 2 3 4 5 6 7 8 9 | log4j.rootLogger=DEBUG,DRFA log4j.threshhold=INFO log4j.appender.DRFA=org.apache.log4j.DailyRollingFileAppender log4j.appender.DRFA.File=/emr/logpusher/log/logpusher.log log4j.appender.DRFA.DatePattern=.yyyy-MM-dd-HH log4j.appender.DRFA.layout=org.apache.log4j.PatternLayout log4j.appender.DRFA.layout.ConversionPattern=%d{ISO8601} %p %t: %m%n log4j.logger.org.apache.commons.httpclient.contrib.ssl.AuthSSLX509TrustManager=INFO log4j.logger.aws157.instancecontroller=DEBUG |
1 2 3 4 | [ec2-user@ip-10-1-2-175 ~]$ sudo service logpusher stop Stopped process in pidfile `/emr/logpusher/run/logpusher.pid' (pid 13677). [ec2-user@ip-10-1-2-175 ~]$ sudo service logpusher status Running [ OK ] |
(Might need to stop Service-nanny before stopping Logpusher, to properly stop/start Logpusher)
DEBUG on Spark classes :
Use the following EMR config to set DEBUG level for relevant class files.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 | [ { "Classification": "hadoop-log4j", "Properties": { "log4j.logger.org.apache.spark.network.crypto" : "DEBUG" } }, { "Classification": "spark-log4j", "Properties": { "log4j.logger.org.apache.spark.network.crypto" : "DEBUG" } } ] |
DEBUG using spark shell:
Execute the following commands after invoking spark-shell to enable DEBUG logging on respective spark classes like Memstore. You can use the same if you want to reduce the amount of logging from INFO (which is default coming from log4j.properties in the spark conf ) to ERROR.
1 2 3 4 5 6 7 8 9 10 11 12 | import org.apache.log4j.{Level, Logger} Logger.getLogger("org.apache.spark.storage.BlockManagerMasterEndpoint").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.storage.BlockManagerInfo").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.storage.BlockManagerMaster").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.storage.MemoryStore").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.scheduler.cluster.YarnClientSchedulerBackend").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.scheduler.cluster.YarnScheduler").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.scheduler.cluster.YarnSchedulerBackend").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.scheduler.DAGScheduler").setLevel(Level.DEBUG) Logger.getLogger("org.apache.spark.scheduler.TaskSetManager").setLevel(Level.DEBUG) // good for progress tracking Logger.getLogger("org.apache.spark.ContextCleaner").setLevel(Level.DEBUG) Logger.getLogger("akka.remote.ReliableDeliverySupervisor").setLevel(Level.DEBUG) |
EMRFS CLI command like EMRFS SYNC :
/etc/hadoop/conf/log4j.properties
1 | hadoop.root.logger=DEBUG,console |
Logs will be on the console out. We might need to redirect to a File or do both.
Enable Debug on Boto3 client :
1 2 3 4 5 6 7 8 9 10 11 12 13 14 | # Create the boto3 client import boto3 import json import logging logging.basicConfig(level=logging.DEBUG) logger = logging.getLogger()
client = boto3.client('sagemaker-runtime')
endpoint_name = "my-model" # Your endpoint name. payload = {"data":"this makes me ", "k": 5} # Payload for inference. response = client.invoke_endpoint(EndpointName=endpoint_name, ContentType='application/json', Body=json.dumps(payload)) response_body = response['Body'] print(response_body.read()) |