Skip to main content

Query HBase tables with Impala

Listen:

As described in other blog posts, Impala uses Hive Metastore Service to query the underlaying data. In this post I use the Hive-HBase handler to connect Hive and HBase and query the data later with Impala. In the past I've written a tutorial (http://mapredit.blogspot.de/2012/12/using-hives-hbase-handler.html) how to connect HBase and Hive, please follow the instructions there.

This approach offers Data Scientists a wide field of work with data stored in HDFS and / or HBase. You will get the possibility to run queries against your stored data independently which technology and database do you use, simply by querying the different data sources in a fast and easy way.

I use the official available census data gathered in 2000 by the US government. The goal is to push this data as CSV into HBase and query this table per Impala. I've made a demonstration script which is available in my git repository.

Demonstration scenario

The dataset looks pretty simple:

cat DEC_00_SF3_P077_with_ann_noheader.csv

8600000US00601,00601,006015-DigitZCTA,0063-DigitZCTA,11102
8600000US00602,00602,006025-DigitZCTA,0063-DigitZCTA,12869
8600000US00603,00603,006035-DigitZCTA,0063-DigitZCTA,12423
8600000US00604,00604,006045-DigitZCTA,0063-DigitZCTA,33548
8600000US00606,00606,006065-DigitZCTA,0063-DigitZCTA,10603

Create the HBase table:

create 'zipcode_hive', 'id', 'zip', 'desc', 'income'

and create an external table in Hive which looks as follows:

CREATE EXTERNAL TABLE ZIPCODE_HBASE (key STRING,zip STRING,desc1 STRING,desc2 STRING,income STRING) ROW FORMAT DELIMITED FIELDS TERMINATED BY ',' STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler' WITH SERDEPROPERTIES ("hbase.columns.mapping" = ":key,zip:zip,desc:desc1,desc:desc2,income:income") TBLPROPERTIES("hbase.table.name" = "zipcode_hive");

Here we map the Hive tables per HBaseStorageHandler to the HBase scheme we've used in the step above.

After these steps are successfully finished, we need to copy the CSV data into HBase. I chose Pig for this task but you can use a translate table in Hive, too.

Here's my Pig script:

cat PopulateData.pig


copyFromLocal DEC_00_SF3_P077_with_ann_noheader.csv ziptest.csv
A = LOAD 'ziptest.csv' USING PigStorage(',') as (id:chararray, zip:chararray, desc1:chararray, desc2:chararray, income:chararray); STORE A INTO 'hbase://zipcode_hive' USING org.apache.pig.backend.hadoop.hbase.HBaseStorage('zip:zip,desc:desc1,desc:desc2,income:income');

The job takes a few seconds and the data is available per HBase:

scan 'zipcode_hive', LIMIT => 2


ROW                                    COLUMN+CELL                                                                                                    
 8600000US00601                        column=desc:desc1, timestamp=1368880594523, value=006015-DigitZCTA                                             
 8600000US00601                        column=desc:desc2, timestamp=1368880594523, value=0063-DigitZCTA                                               
 8600000US00601                        column=income:income, timestamp=1368880594523, value=11102                                                     
 8600000US00601                        column=zip:zip, timestamp=1368880594523, value=00601                                                           
 8600000US00602                        column=desc:desc1, timestamp=1368880594523, value=006025-DigitZCTA                                             
 8600000US00602                        column=desc:desc2, timestamp=1368880594523, value=0063-DigitZCTA                                               
 8600000US00602                        column=income:income, timestamp=1368880594523, value=12869                                                     
 8600000US00602                        column=zip:zip, timestamp=1368880594523, value=00602 

Now we do the same with Impala:

select * from zipcode_hbase limit 4


Using service name 'impala' for kerberos
Connected to hadoop1:21000
Server version: impalad version 1.0 RELEASE (build d1bf0d1dac339af3692ffa17a5e3fdae0aed751f)
Query: select *
from ZIPCODE_HBASE limit 4
Query finished, fetching results ...
+----------------+------------------+----------------+--------+-------+
| key            | desc1            | desc2          | income | zip   |
+----------------+------------------+----------------+--------+-------+
| 8600000US00601 | 006015-DigitZCTA | 0063-DigitZCTA | 11102  | 00601 |
| 8600000US00602 | 006025-DigitZCTA | 0063-DigitZCTA | 12869  | 00602 |
| 8600000US00603 | 006035-DigitZCTA | 0063-DigitZCTA | 12423  | 00603 |
| 8600000US00604 | 006045-DigitZCTA | 0063-DigitZCTA | 33548  | 00604 |
+----------------+------------------+----------------+--------+-------+
Returned 4 row(s) in 0.42s

Another query to get the incomes between 1,000 and 5,000 US$, sorted by income:

select * from zipcode_hbase where income between '1000' and '5000' order by income DESC limit 20;


+----------------+------------------+----------------+--------+-------+
| key            | desc1            | desc2          | income | zip   |
+----------------+------------------+----------------+--------+-------+
| 8600000US64138 | 641385-DigitZCTA | 6413-DigitZCTA | 49995  | 64138 |
| 8600000US12477 | 124775-DigitZCTA | 1243-DigitZCTA | 49993  | 12477 |
| 8600000US33025 | 330255-DigitZCTA | 3303-DigitZCTA | 49991  | 33025 |
| 8600000US44119 | 441195-DigitZCTA | 4413-DigitZCTA | 49988  | 44119 |
| 8600000US34997 | 349975-DigitZCTA | 3493-DigitZCTA | 49982  | 34997 |
| 8600000US70665 | 706655-DigitZCTA | 7063-DigitZCTA | 49981  | 70665 |
| 8600000US28625 | 286255-DigitZCTA | 2863-DigitZCTA | 49981  | 28625 |
| 8600000US76134 | 761345-DigitZCTA | 7613-DigitZCTA | 49979  | 76134 |
| 8600000US44618 | 446185-DigitZCTA | 4463-DigitZCTA | 49978  | 44618 |
| 8600000US65714 | 657145-DigitZCTA | 6573-DigitZCTA | 49978  | 65714 |
| 8600000US77338 | 773385-DigitZCTA | 7733-DigitZCTA | 49976  | 77338 |
| 8600000US14622 | 146225-DigitZCTA | 1463-DigitZCTA | 49972  | 14622 |
| 8600000US84339 | 843395-DigitZCTA | 8433-DigitZCTA | 49972  | 84339 |
| 8600000US85020 | 850205-DigitZCTA | 8503-DigitZCTA | 49967  | 85020 |
| 8600000US64061 | 640615-DigitZCTA | 6403-DigitZCTA | 49964  | 64061 |
| 8600000US97361 | 973615-DigitZCTA | 9733-DigitZCTA | 49961  | 97361 |
| 8600000US30008 | 300085-DigitZCTA | 3003-DigitZCTA | 49960  | 30008 |
| 8600000US48634 | 486345-DigitZCTA | 4863-DigitZCTA | 49958  | 48634 |
| 8600000US47923 | 479235-DigitZCTA | 4793-DigitZCTA | 49946  | 47923 |
| 8600000US46958 | 469585-DigitZCTA | 4693-DigitZCTA | 49946  | 46958 |
+----------------+------------------+----------------+--------+-------+
Returned 20 row(s) in 1.08s

Comments

Popular posts from this blog

Deal with corrupted messages in Apache Kafka

Under some strange circumstances it can happen that a message in a Kafka topic is corrupted. This happens often by using 3rd party frameworks together with Kafka. Additionally, Kafka < 0.9 has no lock at Log.read() at the consumer read level, but has a lock on Log.write(). This can cause a rare race condition, as described in KAKFA-2477 [1]. Probably a log entry looks like: ERROR Error processing message, stopping consumer: (kafka.tools.ConsoleConsumer$) kafka.message.InvalidMessageException: Message is corrupt (stored crc = xxxxxxxxxx, computed crc = yyyyyyyyyy Kafka-Tools Kafka stores the offset of every consumer in Zookeeper. To read out the offsets, Kafka provides handy tools [2]. But also zkCli.sh can be used, at least to display the consumer and the stored offsets. First we need to find the consumer for a topic (> Kafka 0.9): bin/kafka-consumer-groups.sh --zookeeper management01:2181 --describe --group test Prior to Kafka 0.9 the only possibility to get this inform

Hive query shows ERROR "too many counters"

A hive job face the odd " Too many counters:"  like Ended Job = job_xxxxxx with exception 'org.apache.hadoop.mapreduce.counters.LimitExceededException(Too many counters: 201 max=200)' FAILED: Execution Error, return code 1 from org.apache.hadoop.hive.ql.exec.MapRedTask Intercepting System.exit(1) These happens when operators are used in queries ( Hive Operators ). Hive creates 4 counters per operator, max upto 1000, plus a few additional counters like file read/write, partitions and tables. Hence the number of counter required is going to be dependent upon the query.  To avoid such exception, configure " mapreduce.job.counters.max " in mapreduce-site.xml to a value above 1000. Hive will fail when he is hitting the 1k counts, but other MR jobs not. A number around 1120 should be a good choice. Using " EXPLAIN EXTENDED " and " grep -ri operators | wc -l " print out the used numbers of operators. Use this value to tweak the MR s

GPT & GenAI for Startup Storytelling

OpenAI and Bard   are the most used GenAI tools today; the first one has a massive Microsoft investment, and the other one is an experiment from Google. But did you know that you can also use them to optimize and hack your startup?  For startups, creating pitch scripts, sales emails, and elevator pitches with generative AI (GenAI) can help you not only save time but also validate your marketing and wording. Curious? Here are a few prompt hacks for startups to create,improve, and validate buyer personas, your startup's mission/vision statements, and unique selling proposition (USP) definitions. First Step: Introduce yourself and your startup Introduce yourself, your startup, your website, your idea, your position, and in a few words what you are doing to the chatbot: Prompt : I'm NAME and our startup NAME, with website URL, is doing WHATEVER. With PRODUCT NAME, we aim to change or disrupt INDUSTRY. Bard is able to pull information from your website. I'm not sure if ChatGPT