Hive TsFile

Outline

• TsFile-Hive-Connector User Guide
  • About TsFile-Hive-Connector
  • System Requirements
  • Data Type Correspondence
  • Add Dependency For Hive
  • Creating Tsfile-backed Hive tables
  • Querying from Tsfile-backed Hive tables
    • Select Clause Example
    • Aggregate Clause Example
  • What’s Next

TsFile-Hive-Connector User Guide

About TsFile-Hive-Connector

TsFile-Hive-Connector implements the support of Hive for external data sources of Tsfile type. This enables users to operate Tsfile by Hive.

With this connector, you can

• Load a single TsFile, from either the local file system or hdfs, into hive
• Load all files in a specific directory, from either the local file system or hdfs, into hive
• Query the tsfile through HQL.
• As of now, the write operation is not supported in hive-connector. So, insert operation in HQL is not allowed while operating tsfile through hive.

System Requirements

Note: For more information about how to download and use TsFile, please see the following link: https://github.com/apache/incubator-iotdb/tree/master/tsfile.

Data Type Correspondence

Add Dependency For Hive

To use hive-connector in hive, we should add the hive-connector jar into hive.

After downloading the code of iotdb from https://github.com/apache/incubator-iotdb, you can use the command of mvn clean package -pl hive-connector -am -Dmaven.test.skip=true to get a hive-connector-X.X.X-jar-with-dependencies.jar.

Then in hive, use the command of add jar XXX to add the dependency. For example:

hive> add jar /Users/hive/incubator-iotdb/hive-connector/target/hive-connector-0.9.3-jar-with-dependencies.jar;
Added [/Users/hive/incubator-iotdb/hive-connector/target/hive-connector-0.9.3-jar-with-dependencies.jar] to class path
Added resources: [/Users/hive/incubator-iotdb/hive-connector/target/hive-connector-0.9.3-jar-with-dependencies.jar]

Creating Tsfile-backed Hive tables

To create a Tsfile-backed table, specify the serde as org.apache.iotdb.hive.TsFileSerDe, specify the inputformat as org.apache.iotdb.hive.TSFHiveInputFormat, and the outputformat as org.apache.iotdb.hive.TSFHiveOutputFormat.

Also provide a schema which only contains two fields: time_stamp and sensor_id for the table. time_stamp is the time value of the time series and sensor_id is the name of the sensor you want to extract from the tsfile to hive such as sensor_1. The name of the table can be any valid tables names in hive.

Also provide a location from which hive-connector will pull the most current data for the table.

The location must be a specific directory, it can be on your local file system or HDFS if you have set up Hadoop. If it is in your local file system, the location should look like file:///data/data/sequence/root.baic2.WWS.leftfrontdoor/

At last, you should set the device_id in TBLPROPERTIES to the device name you want to analyze.

For example:

CREATE EXTERNAL TABLE IF NOT EXISTS only_sensor_1(
  time_stamp TIMESTAMP,
  sensor_1 BIGINT)
ROW FORMAT SERDE 'org.apache.iotdb.hive.TsFileSerDe'
STORED AS
  INPUTFORMAT 'org.apache.iotdb.hive.TSFHiveInputFormat'
  OUTPUTFORMAT 'org.apache.iotdb.hive.TSFHiveOutputFormat'
LOCATION '/data/data/sequence/root.baic2.WWS.leftfrontdoor/'
TBLPROPERTIES ('device_id'='root.baic2.WWS.leftfrontdoor.plc1');

In this example we’re pulling the data of root.baic2.WWS.leftfrontdoor.plc1.sensor_1 from the directory of /data/data/sequence/root.baic2.WWS.leftfrontdoor/. This table might result in a description as below:

hive> describe only_sensor_1;
OK
time_stamp              timestamp                  from deserializer
sensor_1                bigint                  from deserializer
Time taken: 0.053 seconds, Fetched: 2 row(s)

At this point, the Tsfile-backed table can be worked with in Hive like any other table.

Querying from Tsfile-backed Hive tables

Before we do any queries, we should set the hive.input.format in hive by executing the following command.

hive> set hive.input.format=org.apache.hadoop.hive.ql.io.HiveInputFormat;

Now, we already have an external table named only_sensor_1 in hive. We can use any query operations through HQL to analyse it.

For example:

Select Clause Example

hive> select * from only_sensor_1 limit 10;
OK
1    1000000
2    1000001
3    1000002
4    1000003
5    1000004
6    1000005
7    1000006
8    1000007
9    1000008
10    1000009
Time taken: 1.464 seconds, Fetched: 10 row(s)

Aggregate Clause Example

hive> select count(*) from only_sensor_1;
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 = jackietien_20191016202416_d1e3e233-d367-4453-b39a-2aac9327a3b6
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>
Job running in-process (local Hadoop)
2019-10-16 20:24:18,305 Stage-1 map = 0%,  reduce = 0%
2019-10-16 20:24:27,443 Stage-1 map = 100%,  reduce = 100%
Ended Job = job_local867757288_0002
MapReduce Jobs Launched:
Stage-Stage-1:  HDFS Read: 0 HDFS Write: 0 SUCCESS
Total MapReduce CPU Time Spent: 0 msec
OK
1000000
Time taken: 11.334 seconds, Fetched: 1 row(s)

What’s Next

We’re currently only supporting read operation. Writing tables to Tsfiles is under development.