I've been doing some investigation lately around using Hadoop, Hive, and Pig to do some data transformation. As part of that I've noticed that the schema of data files doesn't seem to attached to files at all. The data files are just flat files (unless using something like a SequenceFile). Each application that wants to work with those files has its own way of representing the schema of those files.
For example, I load a file into the HDFS and want to transform it with Pig. In order to work effectively with it I need to specify the schema of the file when I load the data:
EMP = LOAD 'myfile' using PigStorage() as { first_name: string, last_name: string, deptno: int};
Now, I know that when storing a file using PigStorage, the schema can optionally be written out along side it, but in order to get a file into Pig in the first place it seems like you need to specify a schema.
If I want to work with the same file in Hive, I need to create a table and specify the schema with that too:
CREATE EXTERNAL TABLE EMP ( first_name string
, last_name string
, empno int)
LOCATION 'myfile';
It seems to me like this is extremely fragile. If the file format changes even slightly then the schema must be manually updated in each application. I'm sure I'm being naive but wouldn't it make sense to store the schema with the data file? That way the data is portable between applications and the barrier to using another tool would be lower since you wouldn't need to re-code the schema for each application.
So the question is: Is there a way to specify the schema of a data file in Hadoop/HDFS or do I need to specify the schema for the data file in each application?
It looks like you are looking for Apache Avro. With Avro your schema is embedded in your data, so you can read it without having to worry about schema issues and it makes schema evolution really easy.
The great thing about Avro is that it is completely integrated in Hadoop and you can use it with a lot of Hadoop sub-projects like Pig and Hive.
For example with Pig you could do:
EMP = LOAD 'myfile.avro' using AvroStorage();
I would advise looking at the documentation for AvroStorage for more details.
You can also work with Avro with Hive as described here but I have not used that personally but it should work the same way.
What you need is HCatalog which is
"Apache HCatalog is a table and storage management service for data
created using Apache Hadoop.
This includes:
Providing a shared schema and data type mechanism.
Providing a table abstraction so that users need not be concerned with where or how
their data is stored.
Providing interoperability across data processing tools such as Pig, Map Reduce, and Hive."
You can take a look at the "data flow example" in the docs to see exactly the scenario you are talking about
Apache Zebra seems to be the tool that could provide a common schema definition across mr, pig and hive. It has its own schema store. MR job can use its built in TableStore to write to HDFS.
Related
I have started learning Hadoop.I understood that HDFS provides distributed storage system and Mapreduce is for data processing.Now i ma reading Hadoop ecosystem.
From the definition of Hive, it is a data ware house built on hadoop for providing SQL like interface.
My question is when hadoop provides HDFS which is falut tolerant , distributed then why hive? Does hive replaces HDFS?.
Does hive provide only sql interface or storage also?
Hive does not replace HDFS. Hive provides sql type interface to data that is stored in HDFS. Its basically used for querying and analysis of data that is stored. Hive in a sense actually eliminates a lot of boiler plate code, that you would have to write if you were using mapreduce. for example just think of how you are going to create different types of joins(left, right, bucketed) or group by clause or any other sql clause in mapreduce and you will get your answer (you lines of code will easily scale to 100's ). Hive provides them out-of-the-box. You dont need to write those lengthy programs in mapreduce. Hive already does that for you.
One thing to note is, Hive itself uses Mapreduce behind the scenes. So any group by, count, join is converted to mapreduce jobs only. You can change this though to Tez/Spark.
for your second question, hive does not provide any storage, it just uses a database (derby as default, MySQL would be a good choice if you want to use a different db) as a metastore just to store the metadata related to the tables, partitions, views, buckets etc.. (metadata is like location of tables, type of data stored in tables, partitions info of the tables, created date, modified date etc..) you create with hive.
To answer your question in comment...
Hive can process structured (csv,txt etc) data & semi-structured(xml,json,parquet etc). It cannot process unstructured data like audio, video etc.
Note: Semi structured data can be handled in DDLs and also through spark to be put into Hive.
I encourage you to learn what is external and managed tables in hive too.
Happy learning.
I was trying to use Hive to query the tables I saved using saveAsTable() provided by Spark DataFrame. Everything works well when I query using hiveContext.sql(). However, when I switch to hive and describe the table, it becomes col, array, something like this and is no longer queryable.
Any ideas how to work it through? Is there a reliable way to make Hive understands the metadata defined in spark instead of explicitly defining the schema?
Sometimes I make use of spark to infer schema from the raw data or read schema from certain file formats like parquet so don't want to create these table that could be inferred automatically.
Thanks a lot for any advice!
I have a database on HDFS (Hadoop filesystem) along with a schema file.
I'm trying to connect to this existing database from hive.
Any pointer are really appreciated.
Not sure what you mean by database, but using the External Table feature of Hive, this is fairly easy. You'll need 3 things: a location to the data, and Input(Output)Format to read (write) your data (rows), and potentially a a SerDe to interpret your data (columns). If you need to keep your Hive Schema and external schema in sync, there isn't really a good way to do it out of the box. You'll have to write some custom code that monitors the source schema and modifies the Hive schema on a schema change. Though non-trivial, it's also fairly easy to do this.
I am new to Hadoop. I ran a map reduce on my data and now I want to query it so I can put it into my website. Is Apache Hive the best way to do that? I would greatly appreciate any help.
Keep in mind that Hive is a batch processing system, which under the hoods converts the SQL statements to bunch of MapReduce jobs with stage builds in between. Also, Hive is a high latency system i.e. based on your dataset sizes you are looking at minutes to hours or even days to process a complicated query.
So, if you want to serve the results from your MapReduce job output in your website, its highly recommended you export the results back to a RDBMS using sqoop and then take it from there.
Or, if the data itself is huge and cannot be exported back to RDBMS. Then another option you could think of is using a NoSQL system like HBase.
welcome to Hadoop!
I highly recommend you watch Cloudera Essentials for Apache Hadoop | Chapter 5: The Hadoop Ecosystem and familiarize yourself with the different ways to transfer data inbound and outbound from your HDFS cluster. The video is easy-to-watch and describes advantages / disadvantages to each tool, but this outline should give you the basics of the Hadoop Ecosystem:
Flume - Data integration and import of flat files into HDFS. Designed for asynchronous data streams (e.g., log files). Distributed, scalable, and extensible. Supports various endpoints. Allows preprocessing on data before loading to HDFS.
Sqoop - Bidirectional transfer of structured data (RDBMS) and HDFS. Permits incremental import to HDFS. RDBMS must support JDBC or ODBC.
Hive - SQL-like interface to Hadoop. Requires table structure. JDBC and/or ODBC is required.
Hbase - Allows interactive access of HDFS. Sits on top of HDFS and apply structure to data. Allows for random reads, scales horizontally with cluster. Not a full query language; only permits get/put/scan operations (can be used with Hive and/or Impala). Row-key indexes only on data. Does not use Map Reduce paradigm.
Impala - Similar to Hive, high-performance SQL Engine for querying vast amounts of data stored in HDFS. Does not use Map Reduce. Good alternative to Hive.
Pig - Data flow language for transforming large datasets. Permits schema optionally defined at runtime. PigServer (Java API) permits programmatic access.
Note: I assume the data you are trying to read already exists in HDFS. However, some of the products in the Hadoop ecosystem may be useful for your application or as a general reference, so I included them.
If you're only looking to get data from HDFS then yes, you can do so via Hive.
However, you'll most beneficiate from it if your data are already organized (for instance, in columns).
Lets take an example : your map-reduce job produced a csv file named wordcount.csv and containing two rows : word and count. This csv file is on HDFS.
Let's now suppose you want to know the occurence of the word "gloubiboulga". You can simply achieve this via the following code :
CREATE TABLE data
(
word STRING,
count INT,
text2 STRING
)
ROW FORMAT DELIMITED FIELDS TERMINATED BY ",";
LOAD DATA LOCAL INPATH '/wordcount.csv'
OVERWRITE INTO TABLE data;
select word, count from data where word=="gloubiboulga";
Please note that while this language looks highly like SQL, you'll still have to learn a few things about it.
Im trying to get a clear understanding on HBASE.
Hive:- It just create a Tabular Structure for the Underlying Files in
HDFS. So that we can enable the user to have Querying Abilities on the
HDFS file. Correct me if im wrong here?
Hbase- Again, we have create a Similar table Structure, But bit more
in Structured way( Column Oriented) again over HDFS File system.
aren't they both Same considering the type of job they does. except that Hive runs on Mapredeuce.
Also is that true that we cant create a Hbase table over an Already existing HDFS file?
Hive shares a very similar structures to traditional RDBMS (But Not all), HQL syntax is almost similar to SQL which is good for Database Programmer from learning perspective where as HBase is completely diffrent in the sense that it can be queried only on the basis of its Row Key.
If you want to design a table in RDBMS, you will be following a structured approach in defining columns concentrating more on attributes, while in Hbase the complete design is concentrated around the data, So depending on the type of query to be used we can design a table in Hbase also the columns will be dynamic and will be changing at Runtime (core feature of NoSQL)
You said aren't they both Same considering the type of job they does. except that Hive runs on Mapredeuce .This is not a simple thinking.Because when a hive query is executed, a mapreduce job will be created and triggered.Depending upon data size and complexity it may consume time, since for each mapreduce job, there are some number of steps to do by JobTracker, initializing tasks like maps,combine,shufflesort, reduce etc.
But in case we access HBase, it directly lookup the data they indexed based on specified Scan or Get parameters. Means it just act as a database.
Hive and HBase are completely different things
Hive is a way to create map/reduce jobs for data that resides on HDFS (can be files or HBase)
HBase is an OLTP oriented key-value store that resides on HDFS and can be used in Map/Reduce jobs
In order for Hive to work it holds metadata that maps the HDFS data into tabular data (since SQL works on tables).
I guess it is also important to note that in recent versions Hive is evolving to go beyond a SQL way to write map/reduce jobs and with what HortonWorks calls the "stinger initiative" they have added a dedicated file format (Orc) and import Hive's performance (e.g. with the upcoming Tez execution engine) to deliver SQL on Hadoop (i.e. relatively fast way to run analytics queries for data stored on Hadoop)
Hive:
It's just create a Tabular Structure for the Underlying Files in HDFS. So that we can enable the user to have SQL-like Querying Abilities on existing HDFS files - with typical latency up to minutes.
However, for best performance it's recommended to ETL data into Hive's ORC format.
HBase:
Unlike Hive, HBase is NOT about running SQL queries over existing data in HDFS.
HBase is a strictly-consistent, distributed, low-latency KEY-VALUE STORE.
From The HBase Definitive Guide:
The canonical use case of Bigtable and HBase is the webtable, that is, the web pages
stored while crawling the Internet.
The row key is the reversed URL of the page—for example, org.hbase.www. There is a
column family storing the actual HTML code, the contents family, as well as others
like anchor, which is used to store outgoing links, another one to store inbound links,
and yet another for metadata like language.
Using multiple versions for the contents family allows you to store a few older copies
of the HTML, and is helpful when you want to analyze how often a page changes, for
example. The timestamps used are the actual times when they were fetched from the
crawled website.
The fact that HBase uses HDFS is just an implementation detail: it allows to run HBase on an existing Hadoop cluster, it guarantees redundant storage of data; but it is not a feature in any other sense.
Also is that true that we cant create a Hbase table over an already
existing HDFS file?
No, it's NOT true. Internally HBase stores data in its HFile format.