Use incremental import in sqoop to load data from Oracle (Part II)

The last post discussed the first part of creating incremental import job in sqoop. This post will continue the discussion for the
2nd part of the of the series.

In the last few posts, I discussed the following:
1. Install Cloudera Hadoop Cluster using Cloudera Manager
2. Configurations after CDH Installation
3. Load Data to Hive Table.
4. Import Data to Hive from Oracle Database
5. Export data from Hive table to Oracle Database.
6. Use Impala to query a Hive table
7. Use incremental import in sqoop to load data from Oracle (Part I)

Although we can pass many arguments to execute the sqoop job just what I did in the last post, there is a better way to manage this kind of work by using sqoop job and can do the similar work as showing the picture below.
sqoop_incremental_2

First, let me create a sqoop job using the following command:
sqoop job \
–create student_job \
— import \
–connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
–username wzhou \
–password wzhou \
–table STUDENT \
–incremental append \
–check-column student_id \
–last-value 7 \
-m 4 \
–split-by major

Once the sqoop job is created, I can use the following commands to show, execute or delete a sqoop job.
sqoop job –list
sqoop job –show student_job
sqoop job –exec student_job
sqoop job –delete student_job

Here are the execution results:
[wzhou@vmhost1 ~]$ sqoop job \
> –create student_job \
> — import \
> –connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
> –username wzhou \
> –password wzhou \
> –table STUDENT \
> –incremental append \
> –check-column student_id \
> –last-value 7 \
> -m 4 \
> –split-by major

Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 13:57:34 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
15/09/25 13:57:35 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.

[wzhou@vmhost1 ~]$ sqoop job –list
Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 13:57:47 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
Available jobs:
student_job

[wzhou@vmhost1 ~]$ sqoop job –show student_job
Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 13:58:07 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
Enter password:
Job: student_job
Tool: import
Options:
—————————-
verbose = false
incremental.last.value = 7
db.connect.string = jdbc:oracle:thin:@enkx3-scan:1521:dbm2
codegen.output.delimiters.escape = 0
codegen.output.delimiters.enclose.required = false
codegen.input.delimiters.field = 0
hbase.create.table = false
db.require.password = true
hdfs.append.dir = true
db.table = STUDENT
codegen.input.delimiters.escape = 0
import.fetch.size = null
accumulo.create.table = false
codegen.input.delimiters.enclose.required = false
db.username = wzhou
codegen.output.delimiters.record = 10
import.max.inline.lob.size = 16777216
hbase.bulk.load.enabled = false
hcatalog.create.table = false
db.clear.staging.table = false
incremental.col = student_id
codegen.input.delimiters.record = 0
enable.compression = false
hive.overwrite.table = false
hive.import = false
codegen.input.delimiters.enclose = 0
accumulo.batch.size = 10240000
hive.drop.delims = false
codegen.output.delimiters.enclose = 0
hdfs.delete-target.dir = false
codegen.output.dir = .
codegen.auto.compile.dir = true
relaxed.isolation = false
mapreduce.num.mappers = 4
accumulo.max.latency = 5000
import.direct.split.size = 0
codegen.output.delimiters.field = 44
export.new.update = UpdateOnly
incremental.mode = AppendRows
hdfs.file.format = TextFile
codegen.compile.dir = /tmp/sqoop-wzhou/compile/5c52e95144738b92047bd2e1e37d1b1f
direct.import = false
db.split.column = major
hive.fail.table.exists = false
db.batch = false

[wzhou@vmhost1 ~]$ sqoop job –exec student_job
Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 14:05:19 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
Enter password:
15/09/25 14:05:22 INFO oracle.OraOopManagerFactory: Data Connector for Oracle and Hadoop is disabled.
15/09/25 14:05:22 INFO manager.SqlManager: Using default fetchSize of 1000
15/09/25 14:05:22 INFO tool.CodeGenTool: Beginning code generation
15/09/25 14:05:23 INFO manager.OracleManager: Time zone has been set to GMT
15/09/25 14:05:23 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM STUDENT t WHERE 1=0
15/09/25 14:05:23 INFO orm.CompilationManager: HADOOP_MAPRED_HOME is /usr/lib/hadoop-mapreduce
Note: /tmp/sqoop-wzhou/compile/67a12da72d209152325de07df2841b68/STUDENT.java uses or overrides a deprecated API.
Note: Recompile with -Xlint:deprecation for details.
15/09/25 14:05:25 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-wzhou/compile/67a12da72d209152325de07df2841b68/STUDENT.jar
15/09/25 14:05:26 INFO manager.OracleManager: Time zone has been set to GMT
15/09/25 14:05:26 INFO tool.ImportTool: Maximal id query for free form incremental import: SELECT MAX(student_id) FROM STUDENT
15/09/25 14:05:26 INFO tool.ImportTool: Incremental import based on column student_id
15/09/25 14:05:26 INFO tool.ImportTool: No new rows detected since last import.

As expected, no rows were inserted. So let me insert a few more rows and rerun the jobs.
Run the following SQL commands to insert the rows on Oracle database.
insert into wzhou.student values ( 8, ‘student8’, ‘history’);
insert into wzhou.student values ( 9, ‘student9’, ‘math’);
insert into wzhou.student values ( 10, ‘student10’, ‘computer’ );
commit;

Rerun the job and it shows 3 rows inserted.
[wzhou@vmhost1 ~]$ sqoop job –exec student_job
Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 14:08:49 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
Enter password:
15/09/25 14:08:53 INFO oracle.OraOopManagerFactory: Data Connector for Oracle and Hadoop is disabled.
15/09/25 14:08:53 INFO manager.SqlManager: Using default fetchSize of 1000
15/09/25 14:08:53 INFO tool.CodeGenTool: Beginning code generation
15/09/25 14:08:53 INFO manager.OracleManager: Time zone has been set to GMT
15/09/25 14:08:53 INFO manager.SqlManager: Executing SQL statement: SELECT t.* FROM STUDENT t WHERE 1=0
15/09/25 14:08:53 INFO orm.CompilationManager: HADOOP_MAPRED_HOME is /usr/lib/hadoop-mapreduce
Note: /tmp/sqoop-wzhou/compile/b4ec871b912ecfebe6b03a510f000581/STUDENT.java uses or overrides a deprecated API.
Note: Recompile with -Xlint:deprecation for details.
15/09/25 14:08:55 INFO orm.CompilationManager: Writing jar file: /tmp/sqoop-wzhou/compile/b4ec871b912ecfebe6b03a510f000581/STUDENT.jar
15/09/25 14:08:56 INFO manager.OracleManager: Time zone has been set to GMT
15/09/25 14:08:56 INFO tool.ImportTool: Maximal id query for free form incremental import: SELECT MAX(student_id) FROM STUDENT
15/09/25 14:08:56 INFO tool.ImportTool: Incremental import based on column student_id
15/09/25 14:08:56 INFO tool.ImportTool: Lower bound value: 7
15/09/25 14:08:56 INFO tool.ImportTool: Upper bound value: 10
15/09/25 14:08:56 INFO mapreduce.ImportJobBase: Beginning import of STUDENT
15/09/25 14:08:56 INFO Configuration.deprecation: mapred.jar is deprecated. Instead, use mapreduce.job.jar
15/09/25 14:08:56 INFO Configuration.deprecation: mapred.map.tasks is deprecated. Instead, use mapreduce.job.maps
15/09/25 14:08:56 INFO client.RMProxy: Connecting to ResourceManager at vmhost1.local/192.168.56.71:8032
15/09/25 14:08:59 INFO db.DBInputFormat: Using read commited transaction isolation
15/09/25 14:08:59 INFO db.DataDrivenDBInputFormat: BoundingValsQuery: SELECT MIN(major), MAX(major) FROM STUDENT WHERE ( student_id > 7 AND student_id use test_oracle;
OK
Time taken: 1.548 seconds

hive> select * from student_ext;
OK
2 student2 computer
4 student4 accounting
1 student1 math
3 student3 math
5 student3 computer
7 student5 computer
6 student4 math
10 student10 computer
8 student8 history
9 student9 math

Time taken: 0.694 seconds, Fetched: 10 row(s)

We can see the new rows with student8, 9 and 10.

Use incremental import in sqoop to load data from Oracle (Part I)

This is a two parts series in discussing incremental import job in sqoop. This post is the first part of the series.

In the last few posts, I discussed the following:
1. Install Cloudera Hadoop Cluster using Cloudera Manager
2. Configurations after CDH Installation
3. Load Data to Hive Table.
4. Import Data to Hive from Oracle Database
5. Export data from Hive table to Oracle Database.
6. Use Impala to query a Hive table
When using sqoop to load data to hive table from an Oracle table, it’s not always loading a full table to hive in one shot, just like taking many days’ work to build the house below. In other words, it is common to load partial data from Oracle table to an existing Hive table. This is where we need to use sqoop incremental job to do the work.

sqoop_incremental_1

First, I create a simple table to illustrate the process for incremental import.
1. Create the source table.
Run the following query to create a new table on Oracle database.

<b>create table wzhou.student
(
student_id number(8) not null,
student_name varchar2(20) not null,
major varchar2(20),
CONSTRAINT student_pk PRIMARY KEY (student_id)
);
insert into wzhou.student values ( 1, 'student1', 'math' );
insert into wzhou.student values ( 2, 'student2', 'computer' );
insert into wzhou.student values ( 3, 'student3', 'math' );
insert into wzhou.student values ( 4, 'student4', 'accounting' );

commit;
select * from wzhou.student;</b>

2. Create the import command.

sqoop import \
--connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
--username wzhou \
--password wzhou \
--table STUDENT \
--incremental append \
--check-column student_id \
-m 4 \
--split-by major

check-column argument specifies which column to be checked during the import operation. The column can not be *CHAR type, like VARCHAR2 or CHAR.

incremental argument can have two modes: append and lastmodified. Lasmodified argument is usually used with a lastmodified column defined as timestamp.
Last-value argument is used to specify a value that new rows with greater than this value will be insert.
or use other ways, suggested by one internet article:
–last-value $($HIVE_HOME/bin/hive -e “select max(idcolumn) from tablename”)

Note 1:
If seeing the following error, make sure the tablename specified in –table argument in UPPERCASE. If it is lowercase, you could see the error below.
ERROR tool.ImportTool: Imported Failed: There is no column found in the target table all_objects_inc_test. Please ensure that your table name is correct.

Note 2:
If using –hive-import argument, you could see the following error. It is not supported yet. So have to remove it and build Hive External table after complete the import data to hdfs.
ERROR Append mode for hive imports is not yet supported. Please remove the parameter –append-mode.

3. Execute the Table Import to Hive.
[wzhou@vmhost1 ~]$ sqoop import \
> –connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
> –username wzhou \
> –password wzhou \
> –table STUDENT \
> –incremental append \
> –check-column student_id \
> -m 4 \
> –split-by major

Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 05:14:22 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
15/09/25 05:14:22 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
. . . .
15/09/25 05:17:33 INFO mapreduce.ImportJobBase: Transferred 74 bytes in 184.558 seconds (0.401 bytes/sec)
15/09/25 05:17:33 INFO mapreduce.ImportJobBase: Retrieved 4 records.
15/09/25 05:17:33 INFO util.AppendUtils: Creating missing output directory – STUDENT
15/09/25 05:17:33 INFO tool.ImportTool: Incremental import complete! To run another incremental import of all data following this import, supply the following arguments:
15/09/25 05:17:33 INFO tool.ImportTool: –incremental append
15/09/25 05:17:33 INFO tool.ImportTool: –check-column student_id
15/09/25 05:17:33 INFO tool.ImportTool: –last-value 4
15/09/25 05:17:33 INFO tool.ImportTool: (Consider saving this with ‘sqoop job –create’)

Notice there is a line of –last-value 4 at the end of execution. This is correct as I imported only 4 rows. The result file is under /user/wzhou/STUDENT. Let’s verify it.

[wzhou@vmhost1 ~]$ hdfs dfs -ls /user/wzhou/STUDENT
Found 5 items
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:14 /user/wzhou/STUDENT/part-m-00000
-rw-r--r--   2 wzhou bigdata         42 2015-09-25 05:15 /user/wzhou/STUDENT/part-m-00001
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:16 /user/wzhou/STUDENT/part-m-00002
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00003
-rw-r--r--   2 wzhou bigdata         32 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00004
[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*
2,student2,computer
4,student4,accounting
1,student1,math
3,student3,math
[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*1
2,student2,computer
4,student4,accounting
[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*4
1,student1,math
3,student3,math

Result looks good so far.

4. Create Hive external table. The new hive external table on Hadoop is still under test_oracle database.
hive
USE test_oracle;
CREATE EXTERNAL TABLE student_ext (
student_id string,
student_name string,
major string
)
ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘,’
LINES TERMINATED BY ‘\n’
LOCATION ‘/user/wzhou/STUDENT’;

select * from student_ext;

Here are the result.
[wzhou@vmhost1 ~]$ hive
hive> USE test_oracle;
OK
Time taken: 1.547 seconds
hive> CREATE EXTERNAL TABLE student_ext (
> student_id string,
> student_name string,
> major string
> )
> ROW FORMAT DELIMITED FIELDS TERMINATED BY ‘,’
> LINES TERMINATED BY ‘\n’
> LOCATION ‘/user/wzhou/STUDENT’;

OK
Time taken: 1.266 seconds
hive> select * from student_ext;
OK
2 student2 computer
4 student4 accounting
1 student1 math
3 student3 math
Time taken: 0.679 seconds, Fetched: 4 row(s)

5. 2nd Round of Insert.
Let me to test out more insert to see how incremental import works. On oracle database, create a few more rows.

insert into wzhou.student values ( 5, ‘student3’, ‘computer’);
insert into wzhou.student values ( 6, ‘student4’, ‘math’);
insert into wzhou.student values ( 7, ‘student5’, ‘computer’ );
commit;

6. Rerun the incremental import command.
[wzhou@vmhost1 ~]$ sqoop import \
> –connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
> –username wzhou \
> –password wzhou \
> –table STUDENT \
> –incremental append \
> –check-column student_id \
> -m 4 \
> –split-by major

Warning: /usr/lib/sqoop/../accumulo does not exist! Accumulo imports will fail.
Please set $ACCUMULO_HOME to the root of your Accumulo installation.
15/09/25 05:39:50 INFO sqoop.Sqoop: Running Sqoop version: 1.4.5-cdh5.4.3
15/09/25 05:39:51 WARN tool.BaseSqoopTool: Setting your password on the command-line is insecure. Consider using -P instead.
15/09/25 05:39:51 INFO oracle.OraOopManagerFactory: Data Connector for Oracle and Hadoop is disabled.
. . . .
15/09/25 05:42:54 INFO mapreduce.ImportJobBase: Transferred 130 bytes in 178.3882 seconds (0.7287 bytes/sec)
15/09/25 05:42:54 INFO mapreduce.ImportJobBase: Retrieved 7 records.
15/09/25 05:42:54 INFO util.AppendUtils: Appending to directory STUDENT
15/09/25 05:42:54 INFO util.AppendUtils: Using found partition 5
15/09/25 05:42:54 INFO tool.ImportTool: Incremental import complete! To run another incremental import of all data following this import, supply the following arguments:
15/09/25 05:42:54 INFO tool.ImportTool: –incremental append
15/09/25 05:42:54 INFO tool.ImportTool: –check-column student_id
15/09/25 05:42:54 INFO tool.ImportTool: –last-value 7
15/09/25 05:42:54 INFO tool.ImportTool: (Consider saving this with ‘sqoop job –create’)

Interesting, the result is not what I expect. I expected only three rows will be inserted. But it seem all 7 rows are inserted. Let me verify the result.

[wzhou@vmhost1 ~]$ hdfs dfs -ls /user/wzhou/STUDENT
Found 10 items
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:14 /user/wzhou/STUDENT/part-m-00000
-rw-r--r--   2 wzhou bigdata         42 2015-09-25 05:15 /user/wzhou/STUDENT/part-m-00001
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:16 /user/wzhou/STUDENT/part-m-00002
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00003
-rw-r--r--   2 wzhou bigdata         32 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00004
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:40 /user/wzhou/STUDENT/part-m-00005
-rw-r--r--   2 wzhou bigdata         82 2015-09-25 05:41 /user/wzhou/STUDENT/part-m-00006
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:41 /user/wzhou/STUDENT/part-m-00007
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:42 /user/wzhou/STUDENT/part-m-00008
-rw-r--r--   2 wzhou bigdata         48 2015-09-25 05:42 /user/wzhou/STUDENT/part-m-00009

[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*6
2,student2,computer
4,student4,accounting
5,student3,computer
7,student5,computer

[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*9
1,student1,math
3,student3,math
6,student4,math

Hive result. 
hive> select * from student_ext;
OK
2	student2	computer
4	student4	accounting
1	student1	math
3	student3	math
2	student2	computer
4	student4	accounting
5	student3	computer
7	student5	computer
1	student1	math
3	student3	math
6	student4	math
Time taken: 0.085 seconds, Fetched: 11 row(s)

Ok, let me look at the result from Impala.

[vmhost3:21000] > select count(*) from student_ext;
Query: select count(*) from student_ext
+----------+
| count(*) |
+----------+
| 4        |
+----------+
Fetched 1 row(s) in 0.85s

Interesting. Impala shows 4 rows instead of 11 rows. The reason is that impala does not refresh metadata regularly. So need to do the invalidate metadata to get the last row count. Hive doesn’t seem have this issue.

[vmhost3:21000] > invalidate metadata;
Query: invalidate metadata

Fetched 0 row(s) in 4.48s
[vmhost3:21000] > select count(*) from student_ext;
Query: select count(*) from student_ext
+----------+
| count(*) |
+----------+
| 11       |
+----------+
Fetched 1 row(s) in 0.94s

7. Solution to fix this issue.
The solution is to add one more argument, last-value, to specify where the load stop last time. I redoed the step 1 to 5.
sqoop import \
–connect jdbc:oracle:thin:@enkx3-scan:1521:dbm2 \
–username wzhou \
–password wzhou \
–table STUDENT \
–incremental append \
–check-column student_id \
–last-value 4 \
-m 4 \
–split-by major

Here are the result

HDFS
[wzhou@vmhost1 ~]$ hdfs dfs -ls /user/wzhou/STUDENT
Found 9 items
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:14 /user/wzhou/STUDENT/part-m-00000
-rw-r--r--   2 wzhou bigdata         42 2015-09-25 05:15 /user/wzhou/STUDENT/part-m-00001
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:16 /user/wzhou/STUDENT/part-m-00002
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00003
-rw-r--r--   2 wzhou bigdata         32 2015-09-25 05:17 /user/wzhou/STUDENT/part-m-00004
-rw-r--r--   2 wzhou bigdata         40 2015-09-25 06:02 /user/wzhou/STUDENT/part-m-00005
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 06:02 /user/wzhou/STUDENT/part-m-00006
-rw-r--r--   2 wzhou bigdata          0 2015-09-25 06:03 /user/wzhou/STUDENT/part-m-00007
-rw-r--r--   2 wzhou bigdata         16 2015-09-25 06:04 /user/wzhou/STUDENT/part-m-00008
[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*5
5,student3,computer
7,student5,computer
[wzhou@vmhost1 ~]$ hdfs dfs -cat /user/wzhou/STUDENT/part*8
6,student4,math

Hive
hive> select * from student_ext;
OK
2	student2	computer
4	student4	accounting
1	student1	math
3	student3	math
5	student3	computer
7	student5	computer
6	student4	math
Time taken: 0.095 seconds, Fetched: 7 row(s)

Impala
[vmhost3:21000] > select count(*) from student_ext;
Query: select count(*) from student_ext
+----------+
| count(*) |
+----------+
| 7        |
+----------+
Fetched 1 row(s) in 0.82s

Ok, everything looks good right now.

Use Impala to query a Hive table

Previously, I discussed the followings:
In the last few posts, I discussed the following:
1. Install Cloudera Hadoop Cluster using Cloudera Manager
2. Configurations after CDH Installation
3. Load Data to Hive Table.
4. Import Data to Hive from Oracle Database
5. Export data from Hive table to Oracle Database.
Although Hive is popular in Hadoop world, it has its own drawback, like excessive Map/Reduce operations for certain queries and JVM overhead during Map/Reduce. Impala is designed to improve the query performance accessing data on HDFS. Hive is SQL on Hadoop while Impala is the SQL on HDFS. Hive is using MapReduce job to get the query result while Impala is using the its daemons running on the data nodes to directly access the files on HDFS and don’t use Map/Reduce at all.

There are two ways to use Impala to query tables in Hive. One way is to use command line, Impala Shell. Another one is to use Hue GUI. I am going to show both methods one by one.

Use Impala Shell
Impala Shell is a nice tool similar to SQL Plus to setup database and tables and issue queries. The speed of ad hoc queries is much faster than Hive’s query, especially for queries requiring fast response time. Here are the steps in using Impala shell.

1. Logon as wzhou user and start the impala shell.
[wzhou@vmhost1 ~]$ impala-shell
Starting Impala Shell without Kerberos authentication
Error connecting: TTransportException, Could not connect to vmhost1.local:21000
Welcome to the Impala shell. Press TAB twice to see a list of available commands.
Copyright (c) 2012 Cloudera, Inc. All rights reserved.
(Shell build version: Impala Shell v2.2.0-cdh5.4.3 (517bb0f) built on Wed Jun 24 19:17:40 PDT 2015)
[Not connected] >

Note: The prompt shows Not connected. I need to connect the Impala shell to any Data Node with impalad daemon. My cluster is using vmhost2 and vmhost3 as Data Node. So I pick any one of them, use vmhost2 for this test.
[Not connected] > connect vmhost2;
Connected to vmhost2:21000
Server version: impalad version 2.2.0-cdh5.4.3 RELEASE (build 517bb0f71cd604a00369254ac6d88394df83e0f6)
[vmhost2:21000] >

2. Run some queries. Impala can see the same list of databases and tables like Hive does.

[vmhost2:21000] > show databases;
Query: show databases
+------------------+
| name             |
+------------------+
| _impala_builtins |
| default          |
| test1            |
| test_oracle      |
+------------------+
Fetched 4 row(s) in 0.01s 

[vmhost2:21000] > use test_oracle;
Query: use test_oracle

[vmhost2:21000] > show tables;
Query: show tables
+----------------------+
| name                 |
+----------------------+
| my_all_objects       |
| my_all_objects_sqoop |
+----------------------+
Fetched 2 row(s) in 0.01s

[vmhost2:21000] > select * from my_all_objects_sqoop limit 3;
Query: select * from my_all_objects_sqoop limit 3
+-------+-------------+-----------+-------------+-------------+
| owner | object_name | object_id | object_type | create_date |
+-------+-------------+-----------+-------------+-------------+
| SYS   | I_USER1     | 46        | INDEX       | 2013-03-12  |
| SYS   | I_OBJ#      | 3         | INDEX       | 2013-03-12  |
| SYS   | I_IND1      | 41        | INDEX       | 2013-03-12  |
+-------+-------------+-----------+-------------+-------------+
Fetched 3 row(s) in 0.04s

[vmhost2:21000] > select count(*) from my_all_objects_sqoop;
Query: select count(*) from my_all_objects_sqoop
+----------+
| count(*) |
+----------+
| 22519    |
+----------+
Fetched 1 row(s) in 1.00s

Use Hue Web UI
Another way to use Impala is to run query from Hue UI.

1. From Cloudera Manager screen, click Hue. After Hue screen shows up, click Hue Web UI.
impala_hue_1

2. On Hue home screen, click Query Editors, then choose Impala.
impala_hue_2

3. After Impala Query Editor screen shows up, select test_oracle under DATABASE. Input the following query, then click Execute.
select * from my_all_objects_sqoop limit 3;
impala_hue_3

4. Run another query and check out Explain plan.
impala_hue_4

Run Time Comparison between Hive and Impala
Hive
hive> use test_oracle;
OK
Time taken: 0.534 seconds
hive> show tables;
OK
my_all_objects
my_all_objects_sqoop
Time taken: 0.192 seconds, Fetched: 2 row(s)

hive> select count(*) from my_all_objects;
Query ID = wzhou_20150922112626_efde0c06-2f04-44b2-9bf1-47b31e45de03
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=
In order to limit the maximum number of reducers:
set hive.exec.reducers.max=
In order to set a constant number of reducers:
set mapreduce.job.reduces=
Starting Job = job_1442935988315_0002, Tracking URL = http://vmhost1.local:8088/proxy/application_1442935988315_0002/
Kill Command = /usr/lib/hadoop/bin/hadoop job -kill job_1442935988315_0002
Hadoop job information for Stage-1: number of mappers: 1; number of reducers: 1
2015-09-22 11:26:33,759 Stage-1 map = 0%, reduce = 0%
2015-09-22 11:26:46,563 Stage-1 map = 100%, reduce = 0%, Cumulative CPU 2.25 sec
2015-09-22 11:26:58,035 Stage-1 map = 100%, reduce = 100%, Cumulative CPU 3.97 sec
MapReduce Total cumulative CPU time: 3 seconds 970 msec
Ended Job = job_1442935988315_0002
MapReduce Jobs Launched:
Stage-Stage-1: Map: 1 Reduce: 1 Cumulative CPU: 3.97 sec HDFS Read: 2057365 HDFS Write: 6 SUCCESS
Total MapReduce CPU Time Spent: 3 seconds 970 msec
OK
22782
Time taken: 40.881 seconds, Fetched: 1 row(s)

Impala
Wow, 41 seconds to get a row count of 22,782 by using Hive. That seem excessive on a cluster no other jobs running. Ok, let’s look at Impala’s result.
Note: Impala does not poll frequently for metadata changes. So in case you don’t see the table name after the import, just do the following:
invalidate metadata;
show tables;

[wzhou@vmhost1 ~]$ impala-shell
[Not connected] > connect vmhost2;
Connected to vmhost2:21000
Server version: impalad version 2.2.0-cdh5.4.3 RELEASE (build 517bb0f71cd604a00369254ac6d88394df83e0f6)

[vmhost2:21000] > use test_oracle;
Query: use test_oracle

[vmhost2:21000] > show tables;
Query: show tables
+———————-+
| name |
+———————-+
| my_all_objects |
| my_all_objects_sqoop |
+———————-+
Fetched 2 row(s) in 0.01s

[vmhost2:21000] > select count(*) from my_all_objects;
Query: select count(*) from my_all_objects
+———-+
| count(*) |
+———-+
| 22782 |
+———-+
Fetched 1 row(s) in 0.12s

The above result shows Hive took 41 seconds to get the row count of a table with 22, 782 rows while Impala was significant faster and took 0.12 seconds. I know my cluster is small, not powerful and hive is using Map/Reduce. But getting a total row count of 22,000 needs 45 seconds, it seems too much. On the other hand, Impala’s timing looks more reasonable to me. Obviously Map/Reduce is not my option if I want to run some queries that expect fast response time. But if executing a long running job against a huge dataset, Map/Reduce option might still be on the table if considering job fault tolerance.

Use SQL Developer to Access Hive Table on Hadoop

There are many UI or command-line tool to access Hive data on Hadoop and I am not going to list them one by one. I use SQL Developer a lot in accessing Oracle database and like this powerful tool. That makes me wondering whether I can use SQL Developer to access Hive table on HDFS. After some researches, I did find a way to configure SQL Developer to access Hive table. Here are the steps:

1. Download the JDBC Driver.
Goto Cloudera website, download Hive JDBC Driver at http://www.cloudera.com/content/cloudera/en/downloads/connectors/hive/jdbc/hive-jdbc-v2-5-4.html.
As for now, the latest version of Hive JDBC Connector is v2.5.4. Unzip the zip file to your target directory.

2. Configure SQL Developer
Open SQL Developer, click Preference -> Database -> Third Party JDBC Driver. Click Add Entry to add all of the jar files just unzipped. Close the SQL Developer, reopen it.
1_sql_developer_hive_jdbc_driver

3. Create a Hive Connection
Create a new connection. The port used is HiveSever2 port or the value for hive.server2.thrift.port. In my cluster, it is 10000. I still use test_oracle hive database created in my previous steps. Click Test, it should show Success.
2_sql_developer_hive_connection

4. Access Hive Tables
After click Connect, I can connect to my Hive databases on my Hadoop cluster.
3_sql_developer_hive_table
You can see the view and look to access Hive is very similar the way accessing regular oracle table. Awesome! Here are a few more other screens.
Properties Tab
4_sql_developer_hive_table_properties
DDL Tab
The DDL tab is nice one. The drawback for DDL is that everything is in one line. It would be nice to format the DDL string in a more readable way. Anyway, it is not a big deal.
5_sql_developer_hive_table_ddl
Detail Tab
6_sql_developer_hive_table_detail
Model Tab
7_sql_developer_hive_table_model

Let me run a query and here is the result, exact the look and feel like I use Oracle.
8_sql_developer_hive_table_query

DBMCLI Utility

dbmcli-cellcli
I like good tools. CELLCLI is a useful tools to perform administration works on Exadata Storage Servers. Now, starting Exadata Storage Server Release 12.1.2.1.0, there is another utility, called DBMCLI to configure and monitor Exadata Database Servers. DBMCLI replaces the /opt/oracle.cellos/compmon/exadata_mon_hw_asr.pl Perl script. The usage of DBMCLI is similar to the usage of CELLCLI.

With DBMCLI, we can start/stop services, list alert history, configure SMTP, and monitor hardware components. Here are a few examples executing on our X3 box in the lab. I removed some similar content from the execution result to save space here.

[root@enkx3db01 ~]# imageinfo
Kernel version: 2.6.39-400.248.3.el6uek.x86_64 #1 SMP Wed Mar 11 18:04:34 PDT 2015 x86_64
Image version: 12.1.2.1.1.150316.2
Image activated: 2015-05-13 07:35:16 -0500
Image status: success
System partition on device: /dev/mapper/VGExaDb-LVDbSys2


[root@enkx3db01 ~]# dbmcli
DBMCLI: Release  - Production on Sat Oct 03 07:55:10 CDT 2015

Copyright (c) 2007, 2014, Oracle.  All rights reserved.

DBMCLI> list dbserver
	 enkx3db01	 online

DBMCLI> list dbserver detail
	 name:              	 enkx3db01
	 bbuStatus:         	 normal
	 coreCount:         	 16
	 cpuCount:          	 32
	 diagHistoryDays:   	 7
	 fanCount:          	 16/16
	 fanStatus:         	 normal
	 id:                	 1302FML051
	 interconnectCount: 	 2
	 ipaddress1:        	 192.168.12.1/24
	 kernelVersion:     	 2.6.39-400.248.3.el6uek.x86_64
	 locatorLEDStatus:  	 off
	 makeModel:         	 Oracle Corporation SUN FIRE X4170 M3
	 metricHistoryDays: 	 7
	 msVersion:         	 OSS_12.1.2.1.1_LINUX.X64_150316.2
	 powerCount:        	 2/2
	 powerStatus:       	 normal
	 releaseImageStatus:	 success
	 releaseVersion:    	 12.1.2.1.1.150316.2
	 releaseTrackingBug:	 20240049
	 status:            	 online
	 temperatureReading:	 18.0
	 temperatureStatus: 	 normal
	 upTime:            	 108 days, 16:29
	 msStatus:          	 running
	 rsStatus:          	 running

DBMCLI> list physicaldisk
	 252:0	 NLV9ZD	 normal
	 252:1	 NGXRZF	 normal
	 252:2	 NLV1JD	 normal
	 252:3	 NH06DD	 normal

DBMCLI> list physicaldisk detail
	 name:              	 252:0
	 deviceId:          	 11
	 diskType:          	 HardDisk
	 enclosureDeviceId: 	 252
	 errMediaCount:     	 0
	 errOtherCount:     	 0
	 makeModel:         	 "HITACHI H106030SDSUN300G"
	 physicalFirmware:  	 A3D0
	 physicalInsertTime:	 2015-05-13T07:32:53-05:00
	 physicalInterface: 	 sas
	 physicalSerial:    	 NLV9ZD
	 physicalSize:      	 279.39677238464355G
	 slotNumber:        	 0
	 status:            	 normal

	 name:              	 252:1
	 deviceId:          	 10
	 diskType:          	 HardDisk
	 enclosureDeviceId: 	 252
	 errMediaCount:     	 0
	 errOtherCount:     	 0
	 makeModel:         	 "HITACHI H106030SDSUN300G"
	 physicalFirmware:  	 A3D0
	 physicalInsertTime:	 2015-05-13T07:32:53-05:00
	 physicalInterface: 	 sas
	 physicalSerial:    	 NGXRZF
	 physicalSize:      	 279.39677238464355G
	 slotNumber:        	 1
	 status:            	 normal

	 name:              	 252:2
	 deviceId:          	 9
	 diskType:          	 HardDisk
	 enclosureDeviceId: 	 252
	 errMediaCount:     	 0
	 errOtherCount:     	 0
	 makeModel:         	 "HITACHI H106030SDSUN300G"
	 physicalFirmware:  	 A3D0
	 physicalInsertTime:	 2015-05-13T07:32:53-05:00
	 physicalInterface: 	 sas
	 physicalSerial:    	 NLV1JD
	 physicalSize:      	 279.39677238464355G
	 slotNumber:        	 2
	 status:            	 normal

	 name:              	 252:3
	 deviceId:          	 8
	 diskType:          	 HardDisk
	 enclosureDeviceId: 	 252
	 errMediaCount:     	 0
	 errOtherCount:     	 0
	 makeModel:         	 "HITACHI H106030SDSUN300G"
	 physicalFirmware:  	 A3D0
	 physicalInsertTime:	 2015-05-13T07:32:53-05:00
	 physicalInterface: 	 sas
	 physicalSerial:    	 NH06DD
	 physicalSize:      	 279.39677238464355G
	 slotNumber:        	 3
	 status:            	 normal
	
DBMCLI> list lun 
	 0_0	 0_0	 normal
	 
DBMCLI> list lun detail
	 name:              	 0_0
	 diskType:          	 HardDisk
	 id:                	 0_0
	 lunSize:           	 835.3940000003204G
	 lunUID:            	 0_0
	 raidLevel:         	 5
	 lunWriteCacheMode: 	 "WriteBack, ReadAheadNone, Direct, No Write Cache if Bad BBU"
	 status:            	 normal
	 	 
DBMCLI> list ibport
	 HCA-1:1	 Active
	 HCA-1:2	 Active


DBMCLI> list ibport detail
	 name:              	 HCA-1:1
	 activeSlave:       	 TRUE
	 dataRate:          	 "40 Gbps"
	 hcaFWVersion:      	 2.11.2012
	 id:                	 0x0010e0000128ce65
	 lid:               	 45
	 linkDowned:        	 0
	 linkIntegrityErrs: 	 0
	 linkRecovers:      	 0
	 physLinkState:     	 LinkUp
	 portNumber:        	 1
	 rcvConstraintErrs: 	 0
	 rcvData:           	 1448988236094
	 rcvErrs:           	 0
	 rcvRemotePhysErrs: 	 0
	 status:            	 Active
	 symbolErrs:        	 0
	 vl15Dropped:       	 0
	 xmtConstraintErrs: 	 0
	 xmtData:           	 237626900021
	 xmtDiscards:       	 0

	 name:              	 HCA-1:2
	 activeSlave:       	 FALSE
	 dataRate:          	 "40 Gbps"
	 hcaFWVersion:      	 2.11.2012
	 id:                	 0x0010e0000128ce66
	 lid:               	 46
	 linkDowned:        	 0
	 linkIntegrityErrs: 	 0
	 linkRecovers:      	 0
	 physLinkState:     	 LinkUp
	 portNumber:        	 2
	 rcvConstraintErrs: 	 0
	 rcvData:           	 22172909573667
	 rcvErrs:           	 0
	 rcvRemotePhysErrs: 	 0
	 status:            	 Active
	 symbolErrs:        	 0
	 vl15Dropped:       	 0
	 xmtConstraintErrs: 	 0
	 xmtData:           	 18855706123959
	 xmtDiscards:       	 0

Here are the result from list alerthistory.

DBMCLI> list alerthistory
	 5_1	 2015-06-16T12:39:32-05:00	 critical	 "A power supply component is suspected of causing a fault with a 100 certainty.  Component Name : /SYS/PS1  Fault class    : fault.chassis.power.ext-fail  Fault message  : http://www.sun.com/msg/SPX86-8003-73"
	 5_2	 2015-07-15T15:59:20-05:00	 clear   	 "A power supply component fault has been cleared.  Component Name       : /SYS/PS1  Trap Additional Info : fault.chassis.power.ext-fail"
	 6_1	 2015-06-25T19:36:44-05:00	 critical	 "File system "/" is 80% full, which is above the 80% threshold. Accelerated space reclamation has started.  This alert will be cleared when file system "/" becomes less than 75% full. Top three directories ordered by total space usage are as follows: /opt        : 7.79G /home        : 7.38G /usr        : 3.06G"
	 7_1	 2015-07-06T14:33:51-05:00	 info    	 "The HDD disk controller battery is performing an unscheduled learn cycle. All disk drives have been placed in WriteThrough caching mode. The flash drives are not affected. Battery Serial Number : 10494  Battery Type          : ibbu08  Battery Temperature   : 39 C  Full Charge Capacity  : 1368 mAh  Relative Charge       : 98%  Ambient Temperature   : 19 C"
	 7_2	 2015-07-06T15:45:31-05:00	 clear   	 "All disk drives are in WriteBack caching mode.  Battery Serial Number : 10494  Battery Type          : ibbu08  Battery Temperature   : 39 C  Full Charge Capacity  : 1368 mAh  Relative Charge       : 72%  Ambient Temperature   : 17 C"
	 8_1	 2015-07-15T15:59:35-05:00	 critical	 "A power supply component is suspected of causing a fault with a 100 certainty.  Component Name : /SYS/PS1  Fault class    : fault.chassis.power.ext-fail  Fault message  : http://www.sun.com/msg/SPX86-8003-73"
	 8_2	 2015-07-15T16:01:37-05:00	 clear   	 "A power supply component fault has been cleared.  Component Name       : /SYS/PS1  Trap Additional Info : fault.chassis.power.ext-fail"
	 9_1	 2015-08-04T22:04:16-05:00	 critical	 "File system "/u01" is 80% full, which is above the 80% threshold.  This alert will be cleared when file system "/u01" becomes less than 75% full. Top three directories ordered by total space usage are as follows: /u01/app        : 147.43G /u01/lost+found        : 16K /u01/stage        : 4K"

If want to check out the detail of alert history, run the following.

DBMCLI> list alerthistory detail
	 name:              	 5_1
	 alertDescription:  	 "A power supply component suspected of causing a fault"
	 alertMessage:      	 "A power supply component is suspected of causing a fault with a 100 certainty.  Component Name : /SYS/PS1  Fault class    : fault.chassis.power.ext-fail  Fault message  : http://www.sun.com/msg/SPX86-8003-73"
	 alertSequenceID:   	 5
	 alertShortName:    	 Hardware
	 alertType:         	 Stateful
	 beginTime:         	 2015-06-16T12:39:32-05:00
	 endTime:           	 2015-07-15T15:59:20-05:00
	 examinedBy:        	 
	 metricObjectName:  	 /SYS/PS1_FAULT
	 notificationState: 	 0
	 sequenceBeginTime: 	 2015-06-16T12:39:32-05:00
	 severity:          	 critical
	 alertAction:       	 "For additional information, please refer to http://www.sun.com/msg/SPX86-8003-73"

	 name:              	 5_2
	 alertDescription:  	 "A power supply component fault cleared"
	 alertMessage:      	 "A power supply component fault has been cleared.  Component Name       : /SYS/PS1  Trap Additional Info : fault.chassis.power.ext-fail"
	 alertSequenceID:   	 5
	 alertShortName:    	 Hardware
	 alertType:         	 Stateful
	 beginTime:         	 2015-07-15T15:59:20-05:00
	 endTime:           	 2015-07-15T15:59:20-05:00
	 examinedBy:        	 
	 metricObjectName:  	 /SYS/PS1_FAULT
	 notificationState: 	 0
	 sequenceBeginTime: 	 2015-06-16T12:39:32-05:00
	 severity:          	 clear
	 alertAction:       	 Informational.

	 name:              	 6_1
	 alertDescription:  	 "File system "/" is 80% full"
	 alertMessage:      	 "File system "/" is 80% full, which is above the 80% threshold. Accelerated space reclamation has started.  This alert will be cleared when file system "/" becomes less than 75% full. Top three directories ordered by total space usage are as follows: /opt        : 7.79G /home        : 7.38G /usr        : 3.06G"
	 alertSequenceID:   	 6
	 alertShortName:    	 Software
	 alertType:         	 Stateful
	 beginTime:         	 2015-06-25T19:36:44-05:00
	 examinedBy:        	 
	 metricObjectName:  	 /
	 notificationState: 	 0
	 sequenceBeginTime: 	 2015-06-25T19:36:44-05:00
	 severity:          	 critical
	 alertAction:       	 "MS includes a file deletion policy that is triggered when file system utilitization is high. Deletion of files is triggered when file utilization reaches 80%. For the / file system, 1) files in metric history directory will be deleted using a policy based on the file modification time stamp. Files older than the number of days set by the metricHistoryDays attribute value will be deleted first, then successive deletions will occur for earlier files, down to files with modification time stamps older than or equal to 10 minutes, or until file system utilization is less than 75%. 2) files in the ADR base directory and LOG_HOME directory will be deleted using a policy based on the file modification time stamp. Files older than the number of days set by the diagHistoryDays attribute value will be deleted first, then successive deletions will occur for earlier files, down to files with modification time stamps older than or equal to 10 minutes, or until file system utilization is less than 75%. The renamed alert.log files and ms-odl generation files that are over 5 MB, and older than the successively-shorter age intervals are also deleted. Crash files that are over 5 MB and older than one day will be deleted.Try to delete more recent files, or files not being automatically purged, to free up space if needed."
	 
. . . .

If just want to check out the critcial alerts, run the following.
DBMCLI> describe alerthistory
name
alertDescription
alertMessage
alertSequenceID
alertShortName
alertType
beginTime
endTime
examinedBy modifiable
failedMail
failedSNMP
metricObjectName
metricValue
notificationState
sequenceBeginTime
severity
alertAction

DBMCLI> list alerthistory where alertType=’critical’;

DBMCLI> list alerthistory where severity=’critical’
5_1 2015-06-16T12:39:32-05:00 critical “A power supply component is suspected of causing a fault with a 100 certainty. Component Name : /SYS/PS1 Fault class : fault.chassis.power.ext-fail Fault message : http://www.sun.com/msg/SPX86-8003-73&#8221;
6_1 2015-06-25T19:36:44-05:00 critical “File system “/” is 80% full, which is above the 80% threshold. Accelerated space reclamation has started. This alert will be cleared when file system “/” becomes less than 75% full. Top three directories ordered by total space usage are as follows: /opt : 7.79G /home : 7.38G /usr : 3.06G”
8_1 2015-07-15T15:59:35-05:00 critical “A power supply component is suspected of causing a fault with a 100 certainty. Component Name : /SYS/PS1 Fault class : fault.chassis.power.ext-fail Fault message : http://www.sun.com/msg/SPX86-8003-73&#8221;
9_1 2015-08-04T22:04:16-05:00 critical “File system “/u01” is 80% full, which is above the 80% threshold. This alert will be cleared when file system “/u01″ becomes less than 75% full. Top three directories ordered by total space usage are as follows: /u01/app : 147.43G /u01/lost+found : 16K /u01/stage : 4K”

DBMCLI> list alerthistory where severity=’critical’ and beginTime > ‘2015-08-01T01:00:01-07:00’
9_1 2015-08-04T22:04:16-05:00 critical “File system “/u01” is 80% full, which is above the 80% threshold. This alert will be cleared when file system “/u01″ becomes less than 75% full. Top three directories ordered by total space usage are as follows: /u01/app : 147.43G /u01/lost+found : 16K /u01/stage : 4K”

We can also find out the meric history information.

DBMCLI> list metrichistory
	 DS_TEMP           	 enkx3db01   	 18.0 C      	 2015-10-03T07:00:39-05:00
	 DS_FANS           	 enkx3db01   	 16          	 2015-10-03T07:00:40-05:00
	 DS_BBU_TEMP       	 enkx3db01   	 35.0 C      	 2015-10-03T07:00:41-05:00
	 DS_CPUT           	 enkx3db01   	 2.6 %       	 2015-10-03T07:00:41-05:00
	 DS_CPUT_MS        	 enkx3db01   	 0.0 %       	 2015-10-03T07:00:41-05:00
	 DS_FSUT           	 /           	 87 %        	 2015-10-03T07:00:41-05:00
	 DS_FSUT           	 /boot       	 7 %         	 2015-10-03T07:00:41-05:00
	 DS_FSUT           	 /mnt/oldroot	 80 %        	 2015-10-03T07:00:41-05:00
	 DS_FSUT           	 /u01        	 87 %        	 2015-10-03T07:00:41-05:00
	 DS_MEMUT          	 enkx3db01   	 82 %        	 2015-10-03T07:00:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T07:00:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T07:00:41-05:00
	 DS_RUNQ           	 enkx3db01   	 1.6         	 2015-10-03T07:00:41-05:00
	 DS_SWAP_IN_BY_SEC 	 enkx3db01   	 4.7 KB/sec  	 2015-10-03T07:00:41-05:00
	 DS_SWAP_OUT_BY_SEC	 enkx3db01   	 0.0 KB/sec  	 2015-10-03T07:00:41-05:00
	 DS_SWAP_USAGE     	 enkx3db01   	 2 %         	 2015-10-03T07:00:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T07:00:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T07:00:41-05:00
	 N_HCA_MB_RCV_SEC  	 enkx3db01   	 1.818 MB/sec	 2015-10-03T07:00:42-05:00
	 N_HCA_MB_TRANS_SEC	 enkx3db01   	 2.075 MB/sec	 2015-10-03T07:00:42-05:00
	 N_IB_MB_RCV_SEC   	 HCA-1:1     	 0.233 MB/sec	 2015-10-03T07:00:42-05:00
	 N_IB_MB_RCV_SEC   	 HCA-1:2     	 1.585 MB/sec	 2015-10-03T07:00:42-05:00
	 N_IB_MB_TRANS_SEC 	 HCA-1:1     	 0.216 MB/sec	 2015-10-03T07:00:42-05:00
	 N_IB_MB_TRANS_SEC 	 HCA-1:2     	 1.859 MB/sec	 2015-10-03T07:00:42-05:00
	 N_IB_UTIL_RCV     	 HCA-1:1     	 0.0 %       	 2015-10-03T07:00:42-05:00
	 N_IB_UTIL_RCV     	 HCA-1:2     	 0.0 %       	 2015-10-03T07:00:42-05:00
	 N_IB_UTIL_TRANS   	 HCA-1:1     	 0.0 %       	 2015-10-03T07:00:42-05:00
	 N_IB_UTIL_TRANS   	 HCA-1:2     	 0.1 %       	 2015-10-03T07:00:42-05:00
	 N_NIC_KB_RCV_SEC  	 enkx3db01   	 0.5 KB/sec  	 2015-10-03T07:00:42-05:00
	 N_NIC_KB_TRANS_SEC	 enkx3db01   	 0.4 KB/sec  	 2015-10-03T07:00:42-05:00
	 DS_TEMP           	 enkx3db01   	 18.0 C      	 2015-10-03T07:01:39-05:00
	 DS_FANS           	 enkx3db01   	 16          	 2015-10-03T07:01:40-05:00
	 DS_BBU_TEMP       	 enkx3db01   	 35.0 C      	 2015-10-03T07:01:41-05:00
	 DS_CPUT           	 enkx3db01   	 3.5 %       	 2015-10-03T07:01:41-05:00
	 DS_CPUT_MS        	 enkx3db01   	 0.0 %       	 2015-10-03T07:01:41-05:00
	 DS_FSUT           	 /           	 87 %        	 2015-10-03T07:01:41-05:00
	 DS_FSUT           	 /boot       	 7 %         	 2015-10-03T07:01:41-05:00
	 DS_FSUT           	 /mnt/oldroot	 80 %        	 2015-10-03T07:01:41-05:00
	 DS_FSUT           	 /u01        	 87 %        	 2015-10-03T07:01:41-05:00
	 DS_MEMUT          	 enkx3db01   	 82 %        	 2015-10-03T07:01:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T07:01:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T07:01:41-05:00
	 DS_RUNQ           	 enkx3db01   	 1.3         	 2015-10-03T07:01:41-05:00
	 DS_SWAP_IN_BY_SEC 	 enkx3db01   	 0.0 KB/sec  	 2015-10-03T07:01:41-05:00
	 DS_SWAP_OUT_BY_SEC	 enkx3db01   	 0.0 KB/sec  	 2015-10-03T07:01:41-05:00
	 DS_SWAP_USAGE     	 enkx3db01   	 2 %         	 2015-10-03T07:01:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T07:01:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T07:01:41-05:00
. . . .
	 DS_TEMP           	 enkx3db01   	 18.0 C      	 2015-10-03T08:01:39-05:00
	 DS_FANS           	 enkx3db01   	 16          	 2015-10-03T08:01:40-05:00
	 DS_BBU_TEMP       	 enkx3db01   	 35.0 C      	 2015-10-03T08:01:41-05:00
	 DS_CPUT           	 enkx3db01   	 3.8 %       	 2015-10-03T08:01:41-05:00
	 DS_CPUT_MS        	 enkx3db01   	 0.0 %       	 2015-10-03T08:01:41-05:00
	 DS_FSUT           	 /           	 87 %        	 2015-10-03T08:01:41-05:00
	 DS_FSUT           	 /boot       	 7 %         	 2015-10-03T08:01:41-05:00
	 DS_FSUT           	 /mnt/oldroot	 80 %        	 2015-10-03T08:01:41-05:00
	 DS_FSUT           	 /u01        	 87 %        	 2015-10-03T08:01:41-05:00
	 DS_MEMUT          	 enkx3db01   	 82 %        	 2015-10-03T08:01:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T08:01:41-05:00
	 DS_MEMUT_MS       	 enkx3db01   	 0.4 %       	 2015-10-03T08:01:41-05:00
	 DS_RUNQ           	 enkx3db01   	 1.5         	 2015-10-03T08:01:41-05:00
	 DS_SWAP_IN_BY_SEC 	 enkx3db01   	 0.0 KB/sec  	 2015-10-03T08:01:41-05:00
	 DS_SWAP_OUT_BY_SEC	 enkx3db01   	 0.0 KB/sec  	 2015-10-03T08:01:41-05:00
	 DS_SWAP_USAGE     	 enkx3db01   	 2 %         	 2015-10-03T08:01:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T08:01:41-05:00
	 DS_VIRTMEM_MS     	 enkx3db01   	 4,708 MB    	 2015-10-03T08:01:41-05:00
	 N_HCA_MB_RCV_SEC  	 enkx3db01   	 0.833 MB/sec	 2015-10-03T08:01:43-05:00
	 N_HCA_MB_TRANS_SEC	 enkx3db01   	 0.708 MB/sec	 2015-10-03T08:01:43-05:00
	 N_IB_MB_RCV_SEC   	 HCA-1:1     	 0.095 MB/sec	 2015-10-03T08:01:43-05:00
	 N_IB_MB_RCV_SEC   	 HCA-1:2     	 0.737 MB/sec	 2015-10-03T08:01:43-05:00
	 N_IB_MB_TRANS_SEC 	 HCA-1:1     	 0.081 MB/sec	 2015-10-03T08:01:43-05:00
	 N_IB_MB_TRANS_SEC 	 HCA-1:2     	 0.627 MB/sec	 2015-10-03T08:01:43-05:00
	 N_IB_UTIL_RCV     	 HCA-1:1     	 0.0 %       	 2015-10-03T08:01:43-05:00
	 N_IB_UTIL_RCV     	 HCA-1:2     	 0.0 %       	 2015-10-03T08:01:43-05:00
	 N_IB_UTIL_TRANS   	 HCA-1:1     	 0.0 %       	 2015-10-03T08:01:43-05:00
	 N_IB_UTIL_TRANS   	 HCA-1:2     	 0.0 %       	 2015-10-03T08:01:43-05:00
	 N_NIC_KB_RCV_SEC  	 enkx3db01   	 0.6 KB/sec  	 2015-10-03T08:01:43-05:00
	 N_NIC_KB_TRANS_SEC	 enkx3db01   	 0.4 KB/sec  	 2015-10-03T08:01:43-05:00

If just want to know the average number of processes in the run queue, run the following.

DBMCLI&gt; <b>list metrichistory DS_RUNQ</b>
	 DS_RUNQ	 enkx3db01	 1.5	 2015-10-04T08:00:41-05:00
	 DS_RUNQ	 enkx3db01	 1.3	 2015-10-04T08:01:42-05:00
	 DS_RUNQ	 enkx3db01	 1.3	 2015-10-04T08:02:41-05:00
	 DS_RUNQ	 enkx3db01	 1.6	 2015-10-04T08:03:41-05:00
	 DS_RUNQ	 enkx3db01	 1.6	 2015-10-04T08:04:41-05:00
	 DS_RUNQ	 enkx3db01	 1.7	 2015-10-04T08:05:41-05:00
	 DS_RUNQ	 enkx3db01	 1.8	 2015-10-04T08:06:41-05:00
	 DS_RUNQ	 enkx3db01	 1.4	 2015-10-04T08:07:41-05:00
	 DS_RUNQ	 enkx3db01	 1.3	 2015-10-04T08:08:41-05:00
	 DS_RUNQ	 enkx3db01	 1.7	 2015-10-04T08:09:41-05:00
	 DS_RUNQ	 enkx3db01	 1.6	 2015-10-04T08:10:41-05:00
	 DS_RUNQ	 enkx3db01	 1.6	 2015-10-04T08:11:41-05:00
	 DS_RUNQ	 enkx3db01	 1.7	 2015-10-04T08:12:41-05:00
	 DS_RUNQ	 enkx3db01	 2.2	 2015-10-04T08:13:41-05:00
	 DS_RUNQ	 enkx3db01	 1.8	 2015-10-04T08:14:42-05:00
	 DS_RUNQ	 enkx3db01	 1.6	 2015-10-04T08:15:42-05:00
	 DS_RUNQ	 enkx3db01	 1.4	 2015-10-04T08:16:41-05:00

Similarly we can check out other metric history like
DS_MEMUT: The percentage of total physical memory used on the server.
DS_SWAP_IN_BY_SEC: The number of swap pages read in KB per second.
DS_SWAP_OUT_BY_SEC: The number of swap pages written in KB per second.
DS_SWAP_USAGE: The percentage of swap space used.
N_HCA_MB_RCV_SEC: The number of MB received by the InfiniBand interfaces per second.
N_HCA_MB_TRANS_SEC: The number of MB transmitted by the InfiniBand interfaces per second.
N_NIC_KB_RCV_SEC: The number of KB received by the Ethernet interfaces per second.
N_NIC_KB_TRANS_SEC: The number of KB transmitted by the Ethernet interfaces per second.

Another useful feature is that DBMCLI can configure email notifications for database server. I did not perform the following steps, just use the example from Oracle document.

Configure SMTP on database server
DBMCLI> ALTER DBSERVER smtpServer=’my_mail.example.com’, –
smtpFromAddr=’john.doe@example.com’, –
smtpFrom=’John Doe’, –
smtpToAddr=’jane.smith@example.com’, –
snmpSubscriber=((host=host1),(host=host2)), –
notificationPolicy=’clear’, –
notificationMethod=’mail,snmp’

Validate e-mail on a database server
DBMCLI> ALTER DBSERVER VALIDATE MAIL

Change the email format
DBMCLI> ALTER DBSERVER emailFormat=’text’
DBMCLI> ALTER DBSERVER emailFormat=’html’

To get further detail about DBMCLI, check out the document at http://docs.oracle.com/cd/E50790_01/doc/doc.121/e51951/app_dbmcli.htm#DBMMN22053.