Tagged: BigData Toggle Comment Threads | Keyboard Shortcuts

  • Wang 20:56 on 2019-11-11 Permalink | Reply
    Tags: BigData, , , ,   

    Include Ranger to protect your hadoop ecosystem 

    Apache Ranger

    Apache Ranger™ is a framework to enable, monitor and manage comprehensive data security across the Hadoop platform.

    The vision with Ranger is to provide comprehensive security across the Apache Hadoop ecosystem. With the advent of Apache YARN, the Hadoop platform can now support a true data lake architecture. Enterprises can potentially run multiple workloads, in a multi tenant environment. Data security within Hadoop needs to evolve to support multiple use cases for data access, while also providing a framework for central administration of security policies and monitoring of user access.

     
  • Wang 22:21 on 2018-11-05 Permalink | Reply
    Tags: BigData, , , , , , , ,   

    [Presto] Secure with LDAP 

    For security issue we decided to enable LDAP in presto, to deploy presto into kubernetes cluster we build presto image ourselves which include kerberos authentication and LDAP configurations.

    As you see the image structure, configurations under catalog/etc/hive are very important, please pay attention.

    krb5.conf and xxx.keytab are used to connect to kerberos

    password-authenticator.properties and ldap_server.pem under etc, hive.properties and hive-security.json under catalog are used to connect to LDAP.

    password-authenticator.properties

    password-authenticator.name=ldap
    ldap.url=ldaps://<IP>:<PORT>
    ldap.user-bind-pattern=xxxxxx
    ldap.user-base-dn=xxxxxx
    

    hive.properties

    connector.name=hive-hadoop2
    hive.security=file
    security.config-file=<hive-security.json>
    hive.metastore.authentication.type=KERBEROS
    hive.metastore.uri=thrift://<IP>:<PORT>
    hive.metastore.service.principal=<SERVER-PRINCIPAL>
    hive.metastore.client.principal=<CLIENT-PRINCIPAL>
    hive.metastore.client.keytab=<KEYTAB>
    hive.config.resources=core-site.xml, hdfs-site.xml
    

    hive-security.json

    {
      "schemas": [{
        "user": "user_1",
        "schema": "db_1",
        "owner": false
      }, {
        "user": " ",
        "schema": "db_1",
        "owner": false
      }, {
        "user": "user_2",
        "schema": "db_2",
        "owner": false
      }],
      "tables": [{
        "user": "user_1",
        "schema": "db_1",
        "table": "table_1",
        "privileges": ["SELECT"]
      }, {
        "user": "user_1",
        "schema": "db_1",
        "table": "table_2",
        "privileges": ["SELECT"]
      }, {
        "user": "user_2",
        "schema": "db_1",
        "table": ".*",
        "privileges": ["SELECT"]
      }, {
        "user": "user_2",
        "schema": "db_2",
        "table": "table_1",
        "privileges": ["SELECT"]
      }, {
        "user": "user_2",
        "schema": "db_2",
        "table": "table_2",
        "privileges": ["SELECT"]
      }],
      "sessionProperties": [{
        "allow": false
      }]
    }
    
     
  • Wang 16:56 on 2018-05-02 Permalink | Reply
    Tags: BigData, , , ,   

    [Presto] Connect hive by kerberos 

    For data security, hadoop cluster usually implement different security mechanisms, most commonly used mechanism is kerberos. Recently I tested how to connect hive by kerberos in presto.

    1.Add krb5.conf/keytab/hdfs-site.xml/core-site.xml in every node.

    2.Modify etc/jvm.properties, append -Djava.security.krb5.conf=”krb5.conf location”

    3.Create hive.properties under etc/catalog

    cat << 'EOF' > etc/catalog/hive.properties
    connector.name=hive-hadoop2
    
    hive.metastore.uri=thrift://xxx:9083
    hive.metastore.authentication.type=KERBEROS
    hive.metastore.service.principal=xxx@xxx.com
    hive.metastore.client.principal=xxx@xxx.com
    hive.metastore.client.keytab="keytab location"
    
    hive.config.resources="core-site.xml and hdfs-site.xml" location
    EOF
    

    4.Download hadoop-lzo jar into plugin/hive-hadoop2

    wget http://maven.twttr.com/com/hadoop/gplcompression/hadoop-lzo/0.4.16/hadoop-lzo-0.4.16.jar -O plugin/hive-hadoop2
    

    5.Get principal tgt

    export KRB5_CONFIG="krb5.conf location"
    kinit -kt "keytab location" xxx@xxx.com
    

    6.Restart presto

    bin/launcher restart
    
     
  • Wang 22:15 on 2018-04-14 Permalink | Reply
    Tags: BigData,   

    Hiveserver2 – hive-jdbc version conflict 

    When I tested connecting hive by hiveserver2, I got error:

    Exception in thread "main" java.sql.SQLException: Could not open client transport with JDBC Uri: jdbc:hive2://localhost:10000/test_hive: Could not establish connection to jdbc:hive2://localhost:10000/test_hive: Required field 'client_protocol' is unset! Struct:TOpenSessionReq(client_protocol:null, configuration:{set:hiveconf:hive.server2.thrift.resultset.default.fetch.size=1000, use:database=test_hive})
    at org.apache.hive.jdbc.HiveConnection.(HiveConnection.java:224)
    at org.apache.hive.jdbc.HiveDriver.connect(HiveDriver.java:107)
    at java.sql.DriverManager.getConnection(DriverManager.java:664)
    at java.sql.DriverManager.getConnection(DriverManager.java:247)
    at com.rakuten.dsd.api.cdna.Thrift.main(Thrift.java:17)
    Caused by: java.sql.SQLException: Could not establish connection to jdbc:hive2://localhost:10000/test_hive: Required field 'client_protocol' is unset! Struct:TOpenSessionReq(client_protocol:null, configuration:{set:hiveconf:hive.server2.thrift.resultset.default.fetch.size=1000, use:database=test_hive})
    at org.apache.hive.jdbc.HiveConnection.openSession(HiveConnection.java:699)
    at org.apache.hive.jdbc.HiveConnection.(HiveConnection.java:200)
    ... 4 more
    Caused by: org.apache.thrift.TApplicationException: Required field 'client_protocol' is unset! Struct:TOpenSessionReq(client_protocol:null, configuration:{set:hiveconf:hive.server2.thrift.resultset.default.fetch.size=1000, use:database=test_hive})
    at org.apache.thrift.TApplicationException.read(TApplicationException.java:111)
    at org.apache.thrift.TServiceClient.receiveBase(TServiceClient.java:79)
    at org.apache.hive.service.rpc.thrift.TCLIService$Client.recv_OpenSession(TCLIService.java:168)
    at org.apache.hive.service.rpc.thrift.TCLIService$Client.OpenSession(TCLIService.java:155)
    at org.apache.hive.jdbc.HiveConnection.openSession(HiveConnection.java:680)
    ... 5 more
    
    

    After checking, it’s said that the version of hive-jdbc and hive are conflict, so I changed hive-jdbc version as the same as hive, problem solved.

     
  • Wang 20:23 on 2018-04-11 Permalink | Reply
    Tags: BigData, ,   

    Hackson!

     
  • Wang 20:12 on 2018-03-25 Permalink | Reply
    Tags: Ambari, BigData, , ,   

    [Presto] Integrate with Ambari 

    Days before I have installed presto and ambari separately, officially ambari doesn’t support presto, you have to download ambari-presto-service and configure it yourself if you wanna manage presto on ambari.

    So I tried this.

    1.download hdp yum repository

    wget -nv http://public-repo-1.hortonworks.com/HDP/centos6/2.x/updates/2.6.3.0/hdp.repo -O /etc/yum.repos.d/HDP.repo
    

    2.download ambari-presto-service and configure

    version=`hdp-select status hadoop-client | sed 's/hadoop-client - ([0-9].[0-9]).*/1/'`
    mkdir /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO
    wget https://github.com/prestodb/ambari-presto-service/releases/download/v1.2/ambari-presto-1.2.tar.gz
    tar -xvf ambari-presto-1.2.tar.gz -C /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO
    mv /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO/ambari-presto-1.2/* /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO
    rm -rf /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO/ambari-presto-1.2
    chmod -R +x /var/lib/ambari-server/resources/stacks/HDP/$version/services/PRESTO/*
    

    3.restart ambari-server

    ambari-server restart
    

    4.add presto service on ambari, please configure discovery.uri when you add presto service, e.g. discovery.uri: http://coordinator:8285

    After doing this, you could add catalogs and use presto as query engine.

    I did a simple query comparison between Tez and Presto, if you wanna accurate benchmark result, I think this benchmark test could help. The query is to calculate sum on a hive table.

    Presto: 4s

    presto:test> select sum(count) as sum from (
              -> select count(*) as count from t0004998 where month = '6.5'
              -> union
              -> select count(*) as count from t0004998 where typestatus in ('VL2216','VL2217','VL2218','VL2219','VL2220')
              -> union
              -> select count(*) as count from t0004998 where countrycode in ('FAMILY','FORM','GENUS','KINGDOM','ORDER','PHYLUM','SPECIES')
              -> ) t;
      sum   
    --------
     307374 
    (1 row)
    
    Query 20180317_102034_00040_sq83e, FINISHED, 1 node
    Splits: 29 total, 29 done (100.00%)
    0:04 [982K rows, 374MB] [231K rows/s, 87.8MB/s]
    

    Tez: 29.77s

    hive> select sum(count) from (
        > select count(*) as count from t0004998 where month = "6.5"
        > union
        > select count(*) as count from t0004998 where typestatus in ("VL2216","VL2217","VL2218","VL2219","VL2220")
        > union
        > select count(*) as count from t0004998 where countrycode in ("FAMILY","FORM","GENUS","KINGDOM","ORDER","PHYLUM","SPECIES")
        > ) t;
    Query ID = hdfs_20180317102109_5fd30986-f840-450e-aedd-b51c5e3a48f1
    Total jobs = 1
    Launching Job 1 out of 1
    Status: Running (Executing on YARN cluster with App id application_1521267007048_0012)
    
    --------------------------------------------------------------------------------
            VERTICES      STATUS  TOTAL  COMPLETED  RUNNING  PENDING  FAILED  KILLED
    --------------------------------------------------------------------------------
    Map 1 ..........   SUCCEEDED      1          1        0        0       0       0
    Map 10 .........   SUCCEEDED      1          1        0        0       1       0
    Map 8 ..........   SUCCEEDED      1          1        0        0       0       0
    Reducer 11 .....   SUCCEEDED      1          1        0        0       0       0
    Reducer 2 ......   SUCCEEDED      1          1        0        0       0       1
    Reducer 4 ......   SUCCEEDED      1          1        0        0       0       0
    Reducer 6 ......   SUCCEEDED      1          1        0        0       0       0
    Reducer 7 ......   SUCCEEDED      1          1        0        0       0       0
    Reducer 9 ......   SUCCEEDED      1          1        0        0       0       0
    --------------------------------------------------------------------------------
    VERTICES: 09/09  [==========================>>] 100%  ELAPSED TIME: 29.77 s    
    --------------------------------------------------------------------------------
    OK
    307374
    Time taken: 30.732 seconds, Fetched: 1 row(s)
    
     
  • Wang 21:36 on 2018-03-20 Permalink | Reply
    Tags: BigData, , ,   

    [Presto] Build pseudo cluster 

    Presto is a distributed query engine which is developed by Facebook, for specific concept and advantages, please refer to the official document, below are the steps how I build pseudo cluster on my mac.

    1.download presto

    wget https://repo1.maven.org/maven2/com/facebook/presto/presto-server/0.196/presto-server-0.196.tar.gz
    tar -zvxf presto-server-0.196.tar.gz && cd presto-server-0.196
    

    2.configure configurations

    mkdir etc
    
    cat << 'EOF' > etc/jvm.config
    -server
    -Xmx16G
    -Xms16G
    -XX:+UseG1GC
    -XX:G1HeapRegionSize=32M
    -XX:+UseGCOverheadLimit
    -XX:+ExplicitGCInvokesConcurrent
    -XX:+HeapDumpOnOutOfMemoryError
    -XX:+ExitOnOutOfMemoryError
    EOF
    
    cat << 'EOF' > etc/log.properties
    com.facebook.presto=INFO
    EOF
    
    cat << 'EOF' > etc/config1.properties
    coordinator=true
    node-scheduler.include-coordinator=true
    http-server.http.port=8001
    query.max-memory=24GB
    query.max-memory-per-node=8GB
    discovery-server.enabled=true
    discovery.uri=http://localhost:8001
    EOF
    
    cat << 'EOF' > etc/config2.properties
    coordinator=false
    node-scheduler.include-coordinator=true
    http-server.http.port=8002
    query.max-memory=24GB
    query.max-memory-per-node=8GB
    discovery-server.enabled=true
    discovery.uri=http://localhost:8001
    EOF
    
    cat << 'EOF' > etc/config3.properties
    coordinator=true
    node-scheduler.include-coordinator=true
    http-server.http.port=8003
    query.max-memory=24GB
    query.max-memory-per-node=8GB
    discovery-server.enabled=true
    discovery.uri=http://localhost:8001
    EOF
    
    cat << 'EOF' > etc/node1.properties
    node.environment=test
    node.id=671d18f9-dd0f-412d-b18c-fe6d7989b040
    node.data-dir=/usr/local/Cellar/presto/0.196/data/node1
    EOF
    
    cat << 'EOF' > etc/node2.properties
    node.environment=test
    node.id=e72fdd91-a135-4936-9a3e-f888c5106ed9
    node.data-dir=/usr/local/Cellar/presto/0.196/data/node2
    EOF
    
    cat << 'EOF' > etc/node3.properties
    node.environment=test
    node.id=6ab76715-1812-4093-95cf-1945f4cfefe3
    node.data-dir=/usr/local/Cellar/presto/0.196/data/node3
    EOF
    

    p.s. If you want to restrict operation, please add access-control.properties as below, only permit read operation.

    cat << 'EOF' > etc/access-control.properties
    access-control.name=read-only
    EOF
    

    3.start presto server

    bin/launcher start --config=etc/config1.properties --node-config=etc/node1.properties
    bin/launcher start --config=etc/config2.properties --node-config=etc/node2.properties
    bin/launcher start --config=etc/config3.properties --node-config=etc/node3.properties
    

    4.downlaod cli

    wget https://repo1.maven.org/maven2/com/facebook/presto/presto-cli/0.196/presto-cli-0.196-executable.jar -O bin/presto-cli
    chmod +x bin/presto-cli
    

    5.create catalogs

    cat << 'EOF' > etc/catalog/mysql.properties
    connector.name=mysql
    connection-url=jdbc:mysql://localhost:3306?useSSL=false
    connection-user=presto
    connection-password=presto
    EOF
    
    cat << 'EOF' > etc/catalog/hive.properties
    connector.name=hive-hadoop2
    hive.metastore.uri=thrift://localhost:9083
    EOF
    

    6.connect

    bin/presto-cli --server localhost:8001 --catalog hive
    
    presto> show catalogs;
     Catalog 
    ---------
     hive    
     mysql   
     system  
    (3 rows)
    
    Query 20180318_045410_00013_sq83e, FINISHED, 1 node
    Splits: 1 total, 1 done (100.00%)
    0:00 [0 rows, 0B] [0 rows/s, 0B/s]
    

    Screenshot:


    P.S. If build cluster, pay attention to below items:

    1.node.id in node.properties in every node must be unique in the cluster, you could generate it by uuid/uuidgen.

    2.query.max-memory-per-node in config.properties better to be half of -Xmx in jvm.config.

     
  • Wang 21:33 on 2018-03-11 Permalink | Reply
    Tags: BigData, , , , ,   

    [Sqoop1] Interact MySQL with HDFS/Hive/HBase 

    install sqoop1 on mac

    brew install sqoop
    

    #if you have set env profiles, uncomment profiles in conf/sqoop-env.sh

    1.MySQL -> HDFS

    1.1.import table

    sqoop import --connect jdbc:mysql://localhost/test --direct --username root --P --table t1 --warehouse-dir /mysql/test --fields-terminated-by ','
    

    1.2.import schema

    sqoop import-all-tables --connect jdbc:mysql://localhost/test --direct --username root -P --warehouse-dir /mysql/test --fields-terminated-by ','
    

    2.MySQL -> Hive

    2.1.import definition

    sqoop create-hive-table --connect jdbc:mysql://localhost/test --table t1 --username root --P --hive-database test
    

    2.2.import table

    sqoop import --connect jdbc:mysql://localhost/test --username root --P --table t1 --hive-import --hive-database test --hive-table t1 --fields-terminated-by ','
    

    2.3.import schema

    sqoop import-all-tables --connect jdbc:mysql://localhost/test --username root --P --hive-import --hive-database test --fields-terminated-by ','
    

    3.MySQL -> HBase

    3.1.definition

    sqoop import --connect jdbc:mysql://localhost/test --username root --P --table t1
    

    3.2.import table, need create table in hbase first

    sqoop import --connect jdbc:mysql://localhost/test --username root --P --table t1 --hbase-bulkload --hbase-table test.t1 --column-family basic --fields-terminated-by ','
    

    3.3.import table without creating table in hbase, but pay attention to hbase/sqoop version

    sqoop import --connect jdbc:mysql://localhost/test --username root --P --table t1 --hbase-bulkload --hbase-create-table --hbase-table test.t1 --column-family basic --fields-terminated-by ','
    

    4.HDFS/Hive/HBase -> MySQL

    sqoop export --connect jdbc:mysql://localhost/test --username root --P --table t1 --export-dir /user/hive/warehouse/test.db/t1 --fields-terminated-by ','
    
     
  • Wang 22:21 on 2018-03-09 Permalink | Reply
    Tags: BigData,   

    [HBase] No columns to insert 

    When I load data from hdfs to hbase, I got error:

    Caused by: java.lang.IllegalArgumentException: No columns to insert
        at org.apache.hadoop.hbase.client.HTable.validatePut(HTable.java:1505)
        at org.apache.hadoop.hbase.client.BufferedMutatorImpl.validatePut(BufferedMutatorImpl.java:147)
        at org.apache.hadoop.hbase.client.BufferedMutatorImpl.doMutate(BufferedMutatorImpl.java:134)
        at org.apache.hadoop.hbase.client.BufferedMutatorImpl.mutate(BufferedMutatorImpl.java:98)
        at org.apache.hadoop.hbase.client.HTable.put(HTable.java:1028)
        at org.apache.hadoop.hive.hbase.HiveHBaseTableOutputFormat$MyRecordWriter.write(HiveHBaseTableOutputFormat.java:146)
        at org.apache.hadoop.hive.hbase.HiveHBaseTableOutputFormat$MyRecordWriter.write(HiveHBaseTableOutputFormat.java:117)
        at org.apache.hadoop.hive.ql.io.HivePassThroughRecordWriter.write(HivePassThroughRecordWriter.java:40)
        at org.apache.hadoop.hive.ql.exec.FileSinkOperator.process(FileSinkOperator.java:762)
        at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:897)
        at org.apache.hadoop.hive.ql.exec.SelectOperator.process(SelectOperator.java:95)
        at org.apache.hadoop.hive.ql.exec.Operator.forward(Operator.java:897)
        at org.apache.hadoop.hive.ql.exec.TableScanOperator.process(TableScanOperator.java:130)
        at org.apache.hadoop.hive.ql.exec.MapOperator$MapOpCtx.forward(MapOperator.java:148)
        at org.apache.hadoop.hive.ql.exec.MapOperator.process(MapOperator.java:547)
        ... 9 more
    

    After reading the document, it said that hbase doesn’t support null value, I checked hdfs files, it indeed contained null value in some properties.

    So I modified the data and reloaded to hbase, I didn’t get the error any more.

     
    • Black Hairstyles 04:11 on 2020-11-28 Permalink | Reply

      Thanks , I have just been looking for information about this subject for a while and yours is the best I have came upon so far. But, what about the conclusion? Are you certain about the source?

      Like

    • Hairstyles 10:13 on 2020-12-03 Permalink | Reply

      I am constantly looking online for articles that can assist me. Thank you!

      Like

    • Crave Freebies 06:34 on 2020-12-07 Permalink | Reply

      Magnificent goods from you, man. I’ve understand your stuff previous to and you’re just extremely fantastic. I really like what you have acquired here, certainly like what you are saying and the way in which you say it. You make it enjoyable and you still take care of to keep it sensible. I can not wait to read much more from you. This is really a great site.

      Like

    • I Fashion Styles 19:11 on 2020-12-07 Permalink | Reply

      Have you ever thought about including a little bit more than just your articles? I mean, what you say is important and all. Nevertheless imagine if you added some great pictures or videos to give your posts more, “pop”! Your content is excellent but with pics and clips, this blog could certainly be one of the most beneficial in its niche. Amazing blog!

      Like

  • Wang 20:37 on 2018-03-06 Permalink | Reply
    Tags: BigData, , , ,   

    [Performance Test] MR vs Tez(2) 

    I test the performance of MR vs Tez again on cluster, I created a new table which contains 28,872,974 rows, below are cluster servers:

    Host

    OS

    Memory

    CPU

    Disk

    Region

    master.c.ambari-195807.internal

    CentOS 7

    13 GB

    Intel Ivy Bridge: 2

    200G

    asia-east1-a

    slave1.c.ambari-195807.internal

    CentOS 7

    13 GB

    Intel Ivy Bridge: 2

    200G

    asia-east1-a

    slave2.c.ambari-195807.internal

    CentOS 7

    13 GB

    Intel Ivy Bridge: 2

    200G

    asia-east1-a

    slave3.c.ambari-195807.internal

    CentOS 7

    13 GB

    Intel Ivy Bridge: 2

    200G

    asia-east1-a

    1.MR

    1.1.create table

    hive> CREATE TABLE gbif.gbif_0004998
        > STORED AS ORC
        > TBLPROPERTIES("orc.compress"="snappy")
        > AS SELECT * FROM gbif.gbif_0004998_ori;
    WARNING: Hive-on-MR is deprecated in Hive 2 and may not be available in the future versions. Consider using a different execution engine (i.e. spark, tez) or using Hive 1.X releases.
    Query ID = gizmo_20180225064259_8df29800-b260-48f5-a409-80d6ea5200ad
    Total jobs = 1
    Launching Job 1 out of 1
    Number of reduce tasks is set to 0 since there's no reduce operator
    Starting Job = job_1519536795015_0001, Tracking URL = http://master.c.ambari-195807.internal:8088/proxy/application_1519536795015_0001/
    Kill Command = /opt/apps/hadoop-2.8.3/bin/hadoop job  -kill job_1519536795015_0001
    Hadoop job information for Stage-1: number of mappers: 43; number of reducers: 0
    2018-02-25 06:43:15,110 Stage-1 map = 0%,  reduce = 0%
    2018-02-25 06:44:15,419 Stage-1 map = 0%,  reduce = 0%, Cumulative CPU 231.6 sec
    2018-02-25 06:44:36,386 Stage-1 map = 2%,  reduce = 0%, Cumulative CPU 380.45 sec
    2018-02-25 06:44:37,810 Stage-1 map = 3%,  reduce = 0%, Cumulative CPU 386.09 sec
    2018-02-25 06:44:41,695 Stage-1 map = 5%,  reduce = 0%, Cumulative CPU 422.02 sec
    ...
    ...
    2018-02-25 06:47:36,112 Stage-1 map = 97%,  reduce = 0%, Cumulative CPU 1388.9 sec
    2018-02-25 06:47:38,185 Stage-1 map = 98%,  reduce = 0%, Cumulative CPU 1392.1 sec
    2018-02-25 06:47:45,434 Stage-1 map = 100%,  reduce = 0%, Cumulative CPU 1402.14 sec
    MapReduce Total cumulative CPU time: 23 minutes 22 seconds 140 msec
    Ended Job = job_1519536795015_0001
    Stage-4 is selected by condition resolver.
    Stage-3 is filtered out by condition resolver.
    Stage-5 is filtered out by condition resolver.
    Moving data to directory hdfs://master.c.ambari-195807.internal:9000/user/hive/warehouse/gbif.db/.hive-staging_hive_2018-02-25_06-42-59_672_2925216554228494176-1/-ext-10002
    Moving data to directory hdfs://master.c.ambari-195807.internal:9000/user/hive/warehouse/gbif.db/gbif_0004998
    MapReduce Jobs Launched: 
    Stage-Stage-1: Map: 43   Cumulative CPU: 1402.14 sec   HDFS Read: 11519083564 HDFS Write: 1210708016 SUCCESS
    Total MapReduce CPU Time Spent: 23 minutes 22 seconds 140 msec
    OK
    Time taken: 288.681 seconds
    

    1.2.query by on condition

    hive> select count(*) as total from gbif_0004998 where mediatype = 'STILLIMAGE';
    WARNING: Hive-on-MR is deprecated in Hive 2 and may not be available in the future versions. Consider using a different execution engine (i.e. spark, tez) or using Hive 1.X releases.
    Query ID = gizmo_20180225065438_d2343424-5178-4c44-8b9d-0b28f8b701fa
    Total jobs = 1
    Launching Job 1 out of 1
    Number of reduce tasks determined at compile time: 1
    In order to change the average load for a reducer (in bytes):
      set hive.exec.reducers.bytes.per.reducer=<number>
    In order to limit the maximum number of reducers:
      set hive.exec.reducers.max=<number>
    In order to set a constant number of reducers:
      set mapreduce.job.reduces=<number>
    Starting Job = job_1519536795015_0002, Tracking URL = http://master.c.ambari-195807.internal:8088/proxy/application_1519536795015_0002/
    Kill Command = /opt/apps/hadoop-2.8.3/bin/hadoop job  -kill job_1519536795015_0002
    Hadoop job information for Stage-1: number of mappers: 5; number of reducers: 1
    2018-02-25 06:54:50,078 Stage-1 map = 0%,  reduce = 0%
    2018-02-25 06:55:02,485 Stage-1 map = 40%,  reduce = 0%, Cumulative CPU 21.01 sec
    2018-02-25 06:55:03,544 Stage-1 map = 80%,  reduce = 0%, Cumulative CPU 38.51 sec
    2018-02-25 06:55:06,704 Stage-1 map = 100%,  reduce = 0%, Cumulative CPU 49.23 sec
    2018-02-25 06:55:09,881 Stage-1 map = 100%,  reduce = 100%, Cumulative CPU 51.88 sec
    MapReduce Total cumulative CPU time: 51 seconds 880 msec
    Ended Job = job_1519536795015_0002
    MapReduce Jobs Launched: 
    Stage-Stage-1: Map: 5  Reduce: 1   Cumulative CPU: 51.88 sec   HDFS Read: 1936305 HDFS Write: 107 SUCCESS
    Total MapReduce CPU Time Spent: 51 seconds 880 msec
    OK
    2547716
    Time taken: 32.292 seconds, Fetched: 1 row(s)
    

    1.3.query by two conditions

    hive> select count(*) as total from gbif_0004998 where mediatype = 'STILLIMAGE' and year > 1900;
    WARNING: Hive-on-MR is deprecated in Hive 2 and may not be available in the future versions. Consider using a different execution engine (i.e. spark, tez) or using Hive 1.X releases.
    Query ID = gizmo_20180225081238_766d3707-7eb4-4818-860e-887c48d507ce
    Total jobs = 1
    Launching Job 1 out of 1
    Number of reduce tasks determined at compile time: 1
    In order to change the average load for a reducer (in bytes):
      set hive.exec.reducers.bytes.per.reducer=<number>
    In order to limit the maximum number of reducers:
      set hive.exec.reducers.max=<number>
    In order to set a constant number of reducers:
      set mapreduce.job.reduces=<number>
    Starting Job = job_1519545228015_0002, Tracking URL = http://master.c.ambari-195807.internal:8088/proxy/application_1519545228015_0002/
    Kill Command = /opt/apps/hadoop-2.8.3/bin/hadoop job  -kill job_1519545228015_0002
    Hadoop job information for Stage-1: number of mappers: 5; number of reducers: 1
    2018-02-25 08:17:31,666 Stage-1 map = 0%,  reduce = 0%
    2018-02-25 08:17:43,866 Stage-1 map = 20%,  reduce = 0%, Cumulative CPU 10.58 sec
    2018-02-25 08:17:46,045 Stage-1 map = 60%,  reduce = 0%, Cumulative CPU 34.12 sec
    2018-02-25 08:17:54,996 Stage-1 map = 80%,  reduce = 0%, Cumulative CPU 41.73 sec
    2018-02-25 08:17:57,126 Stage-1 map = 100%,  reduce = 0%, Cumulative CPU 51.37 sec
    2018-02-25 08:17:58,192 Stage-1 map = 100%,  reduce = 100%, Cumulative CPU 53.72 sec
    MapReduce Total cumulative CPU time: 53 seconds 720 msec
    Ended Job = job_1519545228015_0002
    MapReduce Jobs Launched: 
    Stage-Stage-1: Map: 5  Reduce: 1   Cumulative CPU: 53.72 sec   HDFS Read: 8334197 HDFS Write: 107 SUCCESS
    Total MapReduce CPU Time Spent: 53 seconds 720 msec
    OK
    2547716
    Time taken: 321.138 seconds, Fetched: 1 row(s)
    

    2.Tez

    2.1.create table

    hive> CREATE TABLE gbif.gbif_0004998
        > STORED AS ORC
        > TBLPROPERTIES("orc.compress"="snappy")
        > AS SELECT * FROM gbif.gbif_0004998_ori;
    Query ID = gizmo_20180225075657_bae527a7-7cbd-46d9-afbf-70a5adcdee7c
    Total jobs = 1
    Launching Job 1 out of 1
    Status: Running (Executing on YARN cluster with App id application_1519545228015_0001)
    
    ----------------------------------------------------------------------------------------------
            VERTICES      MODE        STATUS  TOTAL  COMPLETED  RUNNING  PENDING  FAILED  KILLED  
    ----------------------------------------------------------------------------------------------
    Map 1 .......... container     SUCCEEDED      1          1        0        0       0       0  
    ----------------------------------------------------------------------------------------------
    VERTICES: 01/01  [==========================>>] 100%  ELAPSED TIME: 639.61 s   
    ----------------------------------------------------------------------------------------------
    Moving data to directory hdfs://master.c.ambari-195807.internal:9000/user/hive/warehouse/gbif.db/gbif_0004998
    OK
    Time taken: 664.817 seconds
    

    2.2.query by one condition

    hive> select count(*) as total from gbif_0004998 where mediatype = 'STILLIMAGE';
    Query ID = gizmo_20180225080856_d1f13489-30b0-4045-bdeb-e3e5e085e736
    Total jobs = 1
    Launching Job 1 out of 1
    Status: Running (Executing on YARN cluster with App id application_1519545228015_0001)
    
    ----------------------------------------------------------------------------------------------
            VERTICES      MODE        STATUS  TOTAL  COMPLETED  RUNNING  PENDING  FAILED  KILLED  
    ----------------------------------------------------------------------------------------------
    Map 1 .......... container     SUCCEEDED      5          5        0        0       0       0  
    Reducer 2 ...... container     SUCCEEDED      1          1        0        0       0       0  
    ----------------------------------------------------------------------------------------------
    VERTICES: 02/02  [==========================>>] 100%  ELAPSED TIME: 17.91 s    
    ----------------------------------------------------------------------------------------------
    OK
    2547716
    Time taken: 19.255 seconds, Fetched: 1 row(s)
    

    2.2.query by two conditions

    hive> select count(*) as total from gbif_0004998 where mediatype = 'STILLIMAGE' and year > 1900;
    Query ID = gizmo_20180225081200_0279f8e6-544b-4573-858b-33f48bf1fa35
    Total jobs = 1
    Launching Job 1 out of 1
    Status: Running (Executing on YARN cluster with App id application_1519545228015_0001)
    
    ----------------------------------------------------------------------------------------------
            VERTICES      MODE        STATUS  TOTAL  COMPLETED  RUNNING  PENDING  FAILED  KILLED  
    ----------------------------------------------------------------------------------------------
    Map 1 .......... container     SUCCEEDED      5          5        0        0       0       0  
    Reducer 2 ...... container     SUCCEEDED      1          1        0        0       0       0  
    ----------------------------------------------------------------------------------------------
    VERTICES: 02/02  [==========================>>] 100%  ELAPSED TIME: 16.96 s    
    ----------------------------------------------------------------------------------------------
    OK
    2547716
    Time taken: 17.635 seconds, Fetched: 1 row(s)
    

    3.Summary

    Rows: 28,872,974

    TypeCreate TableQuery By One ConditionQuery By Two Conditions
    MR288.681s32.292s321.138s
    Tez664.817s19.255s17.635s

    According to the result, MR is quicker than Tez on creation, but slower than Tez on query, along with query condition’s increase, MR’s query performance became worse.

    But why MR is quicker than Tez on creation, currently I don’t know, need to be investigated later.

    Maybe it has relationship with storage, I have checked the filesystem after the two kinds of creation, it’s different. MR has many small files, but Tez has one much bigger file.

    MR generated files

    Tez generated files

     
c
Compose new post
j
Next post/Next comment
k
Previous post/Previous comment
r
Reply
e
Edit
o
Show/Hide comments
t
Go to top
l
Go to login
h
Show/Hide help
shift + esc
Cancel
%d bloggers like this: