User-Defined Functions (UDFs)

User-defined functions (frequently abbreviated as UDFs) let you code your own application logic for processing column values during an Impala query. For example, a UDF could perform calculations using an external math library, combine several column values into one, do geospatial calculations, or other kinds of tests and transformations that are outside the scope of the built-in SQL operators and functions.

You can use UDFs to simplify query logic when producing reports, or to transform data in flexible ways when copying from one table to another with the INSERT ... SELECT syntax.

You might be familiar with this feature from other database products, under names such as stored functions or stored routines.

Impala support for UDFs is available in Impala 1.2 and higher:

  • In Impala 1.1, using UDFs in a query required using the Hive shell. (Because Impala and Hive share the same metastore database, you could switch to Hive to run just those queries requiring UDFs, then switch back to Impala.)
  • Starting in Impala 1.2, Impala can run both high-performance native code UDFs written in C++, and Java-based Hive UDFs that you might already have written.
  • Impala can run scalar UDFs that return a single value for each row of the result set, and user-defined aggregate functions (UDAFs) that return a value based on a set of rows. Currently, Impala does not support user-defined table functions (UDTFs) or window functions.

UDF Concepts

Depending on your use case, you might write all-new functions, reuse Java UDFs that you have already written for Hive, or port Hive Java UDF code to higher-performance native Impala UDFs in C++. You can code either scalar functions for producing results one row at a time, or more complex aggregate functions for doing analysis across. The following sections discuss these different aspects of working with UDFs.

UDFs and UDAFs

Depending on your use case, the user-defined functions (UDFs) you write might accept or produce different numbers of input and output values:

  • The most general kind of user-defined function (the one typically referred to by the abbreviation UDF) takes a single input value and produces a single output value. When used in a query, it is called once for each row in the result set. For example:
    select customer_name, is_frequent_customer(customer_id) from customers;
    select obfuscate(sensitive_column) from sensitive_data;
  • A user-defined aggregate function (UDAF) accepts a group of values and returns a single value. You use UDAFs to summarize and condense sets of rows, in the same style as the built-in COUNT, MAX(), SUM(), and AVG() functions. When called in a query that uses the GROUP BY clause, the function is called once for each combination of GROUP BY values. For example:
    -- Evaluates multiple rows but returns a single value.
    select closest_restaurant(latitude, longitude) from places;
    
    -- Evaluates batches of rows and returns a separate value for each batch.
    select most_profitable_location(store_id, sales, expenses, tax_rate, depreciation) from franchise_data group by year;
  • Currently, Impala does not support other categories of user-defined functions, such as user-defined table functions (UDTFs) or window functions.

Native Impala UDFs

Impala supports UDFs written in C++, in addition to supporting existing Hive UDFs written in Java. Cloudera recommends using C++ UDFs because the compiled native code can yield higher performance, with UDF execution time often 10x faster for a C++ UDF than the equivalent Java UDF.

Using Hive UDFs with Impala

Impala can run Java-based user-defined functions (UDFs), originally written for Hive, with no changes, subject to the following conditions:

  • The parameters and return value must all use scalar data types supported by Impala. For example, complex or nested types are not supported.
  • Hive/Java UDFs must extend org.apache.hadoop.hive.ql.exec.UDF class.
  • Currently, Hive UDFs that accept or return the TIMESTAMP type are not supported.
  • Prior to CDH 5.7 / Impala 2.5 the return type must be a "Writable" type such as Text or IntWritable, rather than a Java primitive type such as String or int. Otherwise, the UDF returns NULL. In CDH 5.7 / Impala 2.5 and higher, this restriction is lifted, and both UDF arguments and return values can be Java primitive types.
  • Hive UDAFs and UDTFs are not supported.
  • Typically, a Java UDF will execute several times slower in Impala than the equivalent native UDF written in C++.
  • In CDH 5.7 / Impala 2.5 and higher, you can transparently call Hive Java UDFs through Impala, or call Impala Java UDFs through Hive. This feature does not apply to built-in Hive functions. Any Impala Java UDFs created with older versions must be re-created using new CREATE FUNCTION syntax, without any signature for arguments or the return value.

To take full advantage of the Impala architecture and performance features, you can also write Impala-specific UDFs in C++.

For background about Java-based Hive UDFs, see the Hive documentation for UDFs. For examples or tutorials for writing such UDFs, search the web for related blog posts.

The ideal way to understand how to reuse Java-based UDFs (originally written for Hive) with Impala is to take some of the Hive built-in functions (implemented as Java UDFs) and take the applicable JAR files through the UDF deployment process for Impala, creating new UDFs with different names:

  1. Take a copy of the Hive JAR file containing the Hive built-in functions. For example, the path might be like /usr/lib/hive/lib/hive-exec-0.10.0-cdh4.2.0.jar, with different version numbers corresponding to your specific level of CDH.
  2. Use jar tf jar_file to see a list of the classes inside the JAR. You will see names like org/apache/hadoop/hive/ql/udf/UDFLower.class and org/apache/hadoop/hive/ql/udf/UDFOPNegative.class. Make a note of the names of the functions you want to experiment with. When you specify the entry points for the Impala CREATE FUNCTION statement, change the slash characters to dots and strip off the .class suffix, for example org.apache.hadoop.hive.ql.udf.UDFLower and org.apache.hadoop.hive.ql.udf.UDFOPNegative.
  3. Copy that file to an HDFS location that Impala can read. (In the examples here, we renamed the file to hive-builtins.jar in HDFS for simplicity.)
  4. For each Java-based UDF that you want to call through Impala, issue a CREATE FUNCTION statement, with a LOCATION clause containing the full HDFS path of the JAR file, and a SYMBOL clause with the fully qualified name of the class, using dots as separators and without the .class extension. Remember that user-defined functions are associated with a particular database, so issue a USE statement for the appropriate database first, or specify the SQL function name as db_name.function_name. Use completely new names for the SQL functions, because Impala UDFs cannot have the same name as Impala built-in functions.
  5. Call the function from your queries, passing arguments of the correct type to match the function signature. These arguments could be references to columns, arithmetic or other kinds of expressions, the results of CAST functions to ensure correct data types, and so on.

Java UDF Example: Reusing lower() Function

For example, the following impala-shell session creates an Impala UDF my_lower() that reuses the Java code for the Hive lower(): built-in function. We cannot call it lower() because Impala does not allow UDFs to have the same name as built-in functions. From SQL, we call the function in a basic way (in a query with no WHERE clause), directly on a column, and on the results of a string expression:

[localhost:21000] > create database udfs;
[localhost:21000] > use udfs;
localhost:21000] > create function lower(string) returns string location '/user/hive/udfs/hive.jar' symbol='org.apache.hadoop.hive.ql.udf.UDFLower';
ERROR: AnalysisException: Function cannot have the same name as a builtin: lower
[localhost:21000] > create function my_lower(string) returns string location '/user/hive/udfs/hive.jar' symbol='org.apache.hadoop.hive.ql.udf.UDFLower';
[localhost:21000] > select my_lower('Some String NOT ALREADY LOWERCASE');
+----------------------------------------------------+
| udfs.my_lower('some string not already lowercase') |
+----------------------------------------------------+
| some string not already lowercase                  |
+----------------------------------------------------+
Returned 1 row(s) in 0.11s
[localhost:21000] > create table t2 (s string);
[localhost:21000] > insert into t2 values ('lower'),('UPPER'),('Init cap'),('CamelCase');
Inserted 4 rows in 2.28s
[localhost:21000] > select * from t2;
+-----------+
| s         |
+-----------+
| lower     |
| UPPER     |
| Init cap  |
| CamelCase |
+-----------+
Returned 4 row(s) in 0.47s
[localhost:21000] > select my_lower(s) from t2;
+------------------+
| udfs.my_lower(s) |
+------------------+
| lower            |
| upper            |
| init cap         |
| camelcase        |
+------------------+
Returned 4 row(s) in 0.54s
[localhost:21000] > select my_lower(concat('ABC ',s,' XYZ')) from t2;
+------------------------------------------+
| udfs.my_lower(concat('abc ', s, ' xyz')) |
+------------------------------------------+
| abc lower xyz                            |
| abc upper xyz                            |
| abc init cap xyz                         |
| abc camelcase xyz                        |
+------------------------------------------+
Returned 4 row(s) in 0.22s

Java UDF Example: Reusing negative() Function

Here is an example that reuses the Hive Java code for the negative() built-in function. This example demonstrates how the data types of the arguments must match precisely with the function signature. At first, we create an Impala SQL function that can only accept an integer argument. Impala cannot find a matching function when the query passes a floating-point argument, although we can call the integer version of the function by casting the argument. Then we overload the same function name to also accept a floating-point argument.

[localhost:21000] > create table t (x int);
[localhost:21000] > insert into t values (1), (2), (4), (100);
Inserted 4 rows in 1.43s
[localhost:21000] > create function my_neg(bigint) returns bigint location '/user/hive/udfs/hive.jar' symbol='org.apache.hadoop.hive.ql.udf.UDFOPNegative';
[localhost:21000] > select my_neg(4);
+----------------+
| udfs.my_neg(4) |
+----------------+
| -4             |
+----------------+
[localhost:21000] > select my_neg(x) from t;
+----------------+
| udfs.my_neg(x) |
+----------------+
| -2             |
| -4             |
| -100           |
+----------------+
Returned 3 row(s) in 0.60s
[localhost:21000] > select my_neg(4.0);
ERROR: AnalysisException: No matching function with signature: udfs.my_neg(FLOAT).
[localhost:21000] > select my_neg(cast(4.0 as int));
+-------------------------------+
| udfs.my_neg(cast(4.0 as int)) |
+-------------------------------+
| -4                            |
+-------------------------------+
Returned 1 row(s) in 0.11s
[localhost:21000] > create function my_neg(double) returns double location '/user/hive/udfs/hive.jar' symbol='org.apache.hadoop.hive.ql.udf.UDFOPNegative';
[localhost:21000] > select my_neg(4.0);
+------------------+
| udfs.my_neg(4.0) |
+------------------+
| -4               |
+------------------+
Returned 1 row(s) in 0.11s

You can find the sample files mentioned here in the Impala github repo.

Runtime Environment for UDFs

By default, Impala copies UDFs into /tmp, and you can configure this location through the --local_library_dir startup flag for the impalad daemon.

Installing the UDF Development Package

To develop UDFs for Impala, download and install the impala-udf-devel package (RHEL-based distributions) or impala-udf-dev (Ubuntu and Debian). This package contains header files, sample source, and build configuration files.

  1. Browse to: https://archive.cloudera.com/cdh5/
  2. Locate the appropriate .repo or list file for your operating system version, such as the .repo file for CDH 5 on RHEL 7.
  3. Use the yum, zypper, or apt-get commands depending on your operating system. For the package name, specify impala-udf-devel (RHEL-based distributions) or impala-udf-dev (Ubuntu and Debian).

When you are ready to start writing your own UDFs, download the sample code and build scripts from the Cloudera sample UDF github. Then see Writing User-Defined Functions (UDFs) for how to code UDFs, and Examples of Creating and Using UDFs for how to build and run UDFs.

Writing User-Defined Functions (UDFs)

Before starting UDF development, make sure to install the development package and download the UDF code samples, as described in Installing the UDF Development Package.

When writing UDFs:

  • Keep in mind the data type differences as you transfer values from the high-level SQL to your lower-level UDF code. For example, in the UDF code you might be much more aware of how many bytes different kinds of integers require.
  • Use best practices for function-oriented programming: choose arguments carefully, avoid side effects, make each function do a single thing, and so on.

Getting Started with UDF Coding

To understand the layout and member variables and functions of the predefined UDF data types, examine the header file /usr/include/impala_udf/udf.h:

// This is the only Impala header required to develop UDFs and UDAs. This header
// contains the types that need to be used and the FunctionContext object. The context
// object serves as the interface object between the UDF/UDA and the impala process. 

For the basic declarations needed to write a scalar UDF, see the header file udf-sample.h within the sample build environment, which defines a simple function named AddUdf():

#ifndef IMPALA_UDF_SAMPLE_UDF_H
#define IMPALA_UDF_SAMPLE_UDF_H

#include <impala_udf/udf.h>

using namespace impala_udf;

IntVal AddUdf(FunctionContext* context, const IntVal& arg1, const IntVal& arg2);

#endif

For sample C++ code for a simple function named AddUdf(), see the source file udf-sample.cc within the sample build environment:

#include "udf-sample.h"

// In this sample we are declaring a UDF that adds two ints and returns an int.
IntVal AddUdf(FunctionContext* context, const IntVal& arg1, const IntVal& arg2) {
  if (arg1.is_null || arg2.is_null) return IntVal::null();
  return IntVal(arg1.val + arg2.val);
}

// Multiple UDFs can be defined in the same file

Data Types for Function Arguments and Return Values

Each value that a user-defined function can accept as an argument or return as a result value must map to a SQL data type that you could specify for a table column.

Currently, Impala UDFs cannot accept arguments or return values of the Impala complex types (STRUCT, ARRAY, or MAP).

Each data type has a corresponding structure defined in the C++ and Java header files, with two member fields and some predefined comparison operators and constructors:

  • is_null indicates whether the value is NULL or not. val holds the actual argument or return value when it is non-NULL.

  • Each struct also defines a null() member function that constructs an instance of the struct with the is_null flag set.

  • The built-in SQL comparison operators and clauses such as <, >=, BETWEEN, and ORDER BY all work automatically based on the SQL return type of each UDF. For example, Impala knows how to evaluate BETWEEN 1 AND udf_returning_int(col1) or ORDER BY udf_returning_string(col2) without you declaring any comparison operators within the UDF itself.

    For convenience within your UDF code, each struct defines == and != operators for comparisons with other structs of the same type. These are for typical C++ comparisons within your own code, not necessarily reproducing SQL semantics. For example, if the is_null flag is set in both structs, they compare as equal. That behavior of null comparisons is different from SQL (where NULL == NULL is NULL rather than true), but more in line with typical C++ behavior.

  • Each kind of struct has one or more constructors that define a filled-in instance of the struct, optionally with default values.

  • Impala cannot process UDFs that accept the composite or nested types as arguments or return them as result values. This limitation applies both to Impala UDFs written in C++ and Java-based Hive UDFs.

  • You can overload functions by creating multiple functions with the same SQL name but different argument types. For overloaded functions, you must use different C++ or Java entry point names in the underlying functions.

The data types defined on the C++ side (in /usr/include/impala_udf/udf.h) are:

  • IntVal represents an INT column.

  • BigIntVal represents a BIGINT column. Even if you do not need the full range of a BIGINT value, it can be useful to code your function arguments as BigIntVal to make it convenient to call the function with different kinds of integer columns and expressions as arguments. Impala automatically casts smaller integer types to larger ones when appropriate, but does not implicitly cast large integer types to smaller ones.

  • SmallIntVal represents a SMALLINT column.

  • TinyIntVal represents a TINYINT column.

  • StringVal represents a STRING column. It has a len field representing the length of the string, and a ptr field pointing to the string data. It has constructors that create a new StringVal struct based on a null-terminated C-style string, or a pointer plus a length; these new structs still refer to the original string data rather than allocating a new buffer for the data. It also has a constructor that takes a pointer to a FunctionContext struct and a length, that does allocate space for a new copy of the string data, for use in UDFs that return string values.

  • BooleanVal represents a BOOLEAN column.

  • FloatVal represents a FLOAT column.

  • DoubleVal represents a DOUBLE column.

  • TimestampVal represents a TIMESTAMP column. It has a date field, a 32-bit integer representing the Gregorian date, that is, the days past the epoch date. It also has a time_of_day field, a 64-bit integer representing the current time of day in nanoseconds.

Variable-Length Argument Lists

UDFs typically take a fixed number of arguments, with each one named explicitly in the signature of your C++ function. Your function can also accept additional optional arguments, all of the same type. For example, you can concatenate two strings, three strings, four strings, and so on. Or you can compare two numbers, three numbers, four numbers, and so on.

To accept a variable-length argument list, code the signature of your function like this:

StringVal Concat(FunctionContext* context, const StringVal& separator,
  int num_var_args, const StringVal* args);

In the CREATE FUNCTION statement, after the type of the first optional argument, include ... to indicate it could be followed by more arguments of the same type. For example, the following function accepts a STRING argument, followed by one or more additional STRING arguments:

[localhost:21000] > create function my_concat(string, string ...) returns string location '/user/test_user/udfs/sample.so' symbol='Concat';

The call from the SQL query must pass at least one argument to the variable-length portion of the argument list.

When Impala calls the function, it fills in the initial set of required arguments, then passes the number of extra arguments and a pointer to the first of those optional arguments.

Handling NULL Values

For correctness, performance, and reliability, it is important for each UDF to handle all situations where any NULL values are passed to your function. For example, when passed a NULL, UDFs typically also return NULL. In an aggregate function, which could be passed a combination of real and NULL values, you might make the final value into a NULL (as in CONCAT()), ignore the NULL value (as in AVG()), or treat it the same as a numeric zero or empty string.

Each parameter type, such as IntVal or StringVal, has an is_null Boolean member. Test this flag immediately for each argument to your function, and if it is set, do not refer to the val field of the argument structure. The val field is undefined when the argument is NULL, so your function could go into an infinite loop or produce incorrect results if you skip the special handling for NULL.

If your function returns NULL when passed a NULL value, or in other cases such as when a search string is not found, you can construct a null instance of the return type by using its null() member function.

Memory Allocation for UDFs

By default, memory allocated within a UDF is deallocated when the function exits, which could be before the query is finished. The input arguments remain allocated for the lifetime of the function, so you can refer to them in the expressions for your return values. If you use temporary variables to construct all-new string values, use the StringVal() constructor that takes an initial FunctionContext* argument followed by a length, and copy the data into the newly allocated memory buffer.

Thread-Safe Work Area for UDFs

One way to improve performance of UDFs is to specify the optional PREPARE_FN and CLOSE_FN clauses on the CREATE FUNCTION statement. The "prepare" function sets up a thread-safe data structure in memory that you can use as a work area. The "close" function deallocates that memory. Each subsequent call to the UDF within the same thread can access that same memory area. There might be several such memory areas allocated on the same host, as UDFs are parallelized using multiple threads.

Within this work area, you can set up predefined lookup tables, or record the results of complex operations on data types such as STRING or TIMESTAMP. Saving the results of previous computations rather than repeating the computation each time is an optimization known as http://en.wikipedia.org/wiki/Memoization. For example, if your UDF performs a regular expression match or date manipulation on a column that repeats the same value over and over, you could store the last-computed value or a hash table of already-computed values, and do a fast lookup to find the result for subsequent iterations of the UDF.

Each such function must have the signature:

void function_name(impala_udf::FunctionContext*, impala_udf::FunctionContext::FunctionScope)

Currently, only THREAD_SCOPE is implemented, not FRAGMENT_SCOPE. See udf.h for details about the scope values.

Error Handling for UDFs

To handle errors in UDFs, you call functions that are members of the initial FunctionContext* argument passed to your function.

A UDF can record one or more warnings, for conditions that indicate minor, recoverable problems that do not cause the query to stop. The signature for this function is:

bool AddWarning(const char* warning_msg);

For a serious problem that requires cancelling the query, a UDF can set an error flag that prevents the query from returning any results. The signature for this function is:

void SetError(const char* error_msg);

Writing User-Defined Aggregate Functions (UDAFs)

User-defined aggregate functions (UDAFs or UDAs) are a powerful and flexible category of user-defined functions. If a query processes N rows, calling a UDAF during the query condenses the result set, anywhere from a single value (such as with the SUM or MAX functions), or some number less than or equal to N (as in queries using the GROUP BY or HAVING clause).

The Underlying Functions for a UDA

A UDAF must maintain a state value across subsequent calls, so that it can accumulate a result across a set of calls, rather than derive it purely from one set of arguments. For that reason, a UDAF is represented by multiple underlying functions:

  • An initialization function that sets any counters to zero, creates empty buffers, and does any other one-time setup for a query.
  • An update function that processes the arguments for each row in the query result set and accumulates an intermediate result for each node. For example, this function might increment a counter, append to a string buffer, or set flags.
  • A merge function that combines the intermediate results from two different nodes.
  • A serialize function that flattens any intermediate values containing pointers, and frees any memory allocated during the init, update, and merge phases.
  • A finalize function that either passes through the combined result unchanged, or does one final transformation.

In the SQL syntax, you create a UDAF by using the statement CREATE AGGREGATE FUNCTION. You specify the entry points of the underlying C++ functions using the clauses INIT_FN, UPDATE_FN, MERGE_FN, SERIALIZE_FN, and FINALIZE_FN.

For convenience, you can use a naming convention for the underlying functions and Impala automatically recognizes those entry points. Specify the UPDATE_FN clause, using an entry point name containing the string update or Update. When you omit the other _FN clauses from the SQL statement, Impala looks for entry points with names formed by substituting the update or Update portion of the specified name.

uda-sample.h:

See this file online at: uda-sample.h

uda-sample.cc:

See this file online at: uda-sample.cc

Intermediate Results for UDAs

A user-defined aggregate function might produce and combine intermediate results during some phases of processing, using a different data type than the final return value. For example, if you implement a function similar to the built-in AVG() function, it must keep track of two values, the number of values counted and the sum of those values. Or, you might accumulate a string value over the course of a UDA, then in the end return a numeric or Boolean result.

In such a case, specify the data type of the intermediate results using the optional INTERMEDIATE type_name clause of the CREATE AGGREGATE FUNCTION statement. If the intermediate data is a typeless byte array (for example, to represent a C++ struct or array), specify the type name as CHAR(n), with n representing the number of bytes in the intermediate result buffer.

For an example of this technique, see the trunc_sum() aggregate function, which accumulates intermediate results of type DOUBLE and returns BIGINT at the end. View the appropriate CREATE FUNCTION statement and the implementation of the underlying TruncSum*() functions on Github.

Building and Deploying UDFs

This section explains the steps to compile Impala UDFs from C++ source code, and deploy the resulting libraries for use in Impala queries.

Impala UDF development package ships with a sample build environment for UDFs, that you can study, experiment with, and adapt for your own use.

To build the sample environment:
  1. Install the Impala UDF development package as described in Installing the UDF Development Package
  2. Run the following commands:
    cmake .
    make

The cmake configuration command reads the file CMakeLists.txt and generates a Makefile customized for your particular directory paths. Then the make command runs the actual build steps based on the rules in the Makefile.

Impala loads the shared library from an HDFS location. After building a shared library containing one or more UDFs, use hdfs dfs or hadoop fs commands to copy the binary file to an HDFS location readable by Impala.

The final step in deployment is to issue a CREATE FUNCTION statement in the impala-shell interpreter to make Impala aware of the new function. See CREATE FUNCTION Statement for syntax details. Because each function is associated with a particular database, always issue a USE statement to the appropriate database before creating a function, or specify a fully qualified name, that is, CREATE FUNCTION db_name.function_name.

As you update the UDF code and redeploy updated versions of a shared library, use DROP FUNCTION and CREATE FUNCTION to let Impala pick up the latest version of the code.

Prerequisites for the build environment are:

# Use the appropriate package installation command for your Linux distribution.
sudo yum install gcc-c++ cmake boost-devel
sudo yum install impala-udf-devel
# The package name on Ubuntu and Debian is impala-udf-dev.

Then, unpack the sample code in udf_samples.tar.gz and use that as a template to set up your build environment.

To build the original samples:

# Process CMakeLists.txt and set up appropriate Makefiles.
cmake .
# Generate shared libraries from UDF and UDAF sample code,
# udf_samples/libudfsample.so and udf_samples/libudasample.so
make

The sample code to examine, experiment with, and adapt is in these files:

  • udf-sample.h: Header file that declares the signature for a scalar UDF (AddUDF).
  • udf-sample.cc: Sample source for a simple UDF that adds two integers. Because Impala can reference multiple function entry points from the same shared library, you could add other UDF functions in this file and add their signatures to the corresponding header file.
  • udf-sample-test.cc: Basic unit tests for the sample UDF.
  • uda-sample.h: Header file that declares the signature for sample aggregate functions. The SQL functions will be called COUNT, AVG, and STRINGCONCAT. Because aggregate functions require more elaborate coding to handle the processing for multiple phases, there are several underlying C++ functions such as CountInit, AvgUpdate, and StringConcatFinalize.
  • uda-sample.cc: Sample source for simple UDAFs that demonstrate how to manage the state transitions as the underlying functions are called during the different phases of query processing.
    • The UDAF that imitates the COUNT function keeps track of a single incrementing number; the merge functions combine the intermediate count values from each Impala node, and the combined number is returned verbatim by the finalize function.
    • The UDAF that imitates the AVG function keeps track of two numbers, a count of rows processed and the sum of values for a column. These numbers are updated and merged as with COUNT, then the finalize function divides them to produce and return the final average value.
    • The UDAF that concatenates string values into a comma-separated list demonstrates how to manage storage for a string that increases in length as the function is called for multiple rows.
  • uda-sample-test.cc: basic unit tests for the sample UDAFs.

Performance Considerations for UDFs

Because a UDF typically processes each row of a table, potentially being called billions of times, the performance of each UDF is a critical factor in the speed of the overall ETL or ELT pipeline. Tiny optimizations you can make within the function body can pay off in a big way when the function is called over and over when processing a huge result set.

Examples of Creating and Using UDFs

This section demonstrates how to create and use all kinds of user-defined functions (UDFs).

For downloadable examples that you can experiment with, adapt, and use as templates for your own functions, see the Cloudera sample UDF github. You must have already installed the appropriate header files, as explained in Installing the UDF Development Package.

Sample C++ UDFs: HasVowels, CountVowels, StripVowels

This example shows 3 separate UDFs that operate on strings and return different data types. In the C++ code, the functions are HasVowels() (checks if a string contains any vowels), CountVowels() (returns the number of vowels in a string), and StripVowels() (returns a new string with vowels removed).

First, we add the signatures for these functions to udf-sample.h in the demo build environment:

BooleanVal HasVowels(FunctionContext* context, const StringVal& input);
IntVal CountVowels(FunctionContext* context, const StringVal& arg1);
StringVal StripVowels(FunctionContext* context, const StringVal& arg1);

Then, we add the bodies of these functions to udf-sample.cc:

BooleanVal HasVowels(FunctionContext* context, const StringVal& input)
{
        if (input.is_null) return BooleanVal::null();

        int index;
        uint8_t *ptr;

        for (ptr = input.ptr, index = 0; index <= input.len; index++, ptr++)
        {
                uint8_t c = tolower(*ptr);
                if (c == 'a' || c == 'e' || c == 'i' || c == 'o' || c == 'u')
                {
                        return BooleanVal(true);
                }
        }
        return BooleanVal(false);
}

IntVal CountVowels(FunctionContext* context, const StringVal& arg1)
{
        if (arg1.is_null) return IntVal::null();

        int count;
        int index;
        uint8_t *ptr;

        for (ptr = arg1.ptr, count = 0, index = 0; index <= arg1.len; index++, ptr++)
        {
                uint8_t c = tolower(*ptr);
                if (c == 'a' || c == 'e' || c == 'i' || c == 'o' || c == 'u')
                {
                        count++;
                }
        }
        return IntVal(count);
}

StringVal StripVowels(FunctionContext* context, const StringVal& arg1)
{
        if (arg1.is_null) return StringVal::null();

        int index;
        std::string original((const char *)arg1.ptr,arg1.len);
        std::string shorter("");

        for (index = 0; index < original.length(); index++)
        {
                uint8_t c = original[index];
                uint8_t l = tolower(c);

                if (l == 'a' || l == 'e' || l == 'i' || l == 'o' || l == 'u')
                {
                        ;
                }
                else
                {
                    shorter.append(1, (char)c);
                }
        }
// The modified string is stored in 'shorter', which is destroyed when this function ends. We need to make a string val
// and copy the contents.
        StringVal result(context, shorter.size()); // Only the version of the ctor that takes a context object allocates new memory
        memcpy(result.ptr, shorter.c_str(), shorter.size());
        return result;
}

We build a shared library, libudfsample.so, and put the library file into HDFS where Impala can read it:

$ make
[  0%] Generating udf_samples/uda-sample.ll
[ 16%] Built target uda-sample-ir
[ 33%] Built target udasample
[ 50%] Built target uda-sample-test
[ 50%] Generating udf_samples/udf-sample.ll
[ 66%] Built target udf-sample-ir
Scanning dependencies of target udfsample
[ 83%] Building CXX object CMakeFiles/udfsample.dir/udf-sample.o
Linking CXX shared library udf_samples/libudfsample.so
[ 83%] Built target udfsample
Linking CXX executable udf_samples/udf-sample-test
[100%] Built target udf-sample-test
$ hdfs dfs -put ./udf_samples/libudfsample.so /user/hive/udfs/libudfsample.so

Finally, we go into the impala-shell interpreter where we set up some sample data, issue CREATE FUNCTION statements to set up the SQL function names, and call the functions in some queries:

[localhost:21000] > create database udf_testing;
[localhost:21000] > use udf_testing;

[localhost:21000] > create function has_vowels (string) returns boolean location '/user/hive/udfs/libudfsample.so' symbol='HasVowels';
[localhost:21000] > select has_vowels('abc');
+------------------------+
| udfs.has_vowels('abc') |
+------------------------+
| true                   |
+------------------------+
Returned 1 row(s) in 0.13s
[localhost:21000] > select has_vowels('zxcvbnm');
+----------------------------+
| udfs.has_vowels('zxcvbnm') |
+----------------------------+
| false                      |
+----------------------------+
Returned 1 row(s) in 0.12s
[localhost:21000] > select has_vowels(null);
+-----------------------+
| udfs.has_vowels(null) |
+-----------------------+
| NULL                  |
+-----------------------+
Returned 1 row(s) in 0.11s
[localhost:21000] > select s, has_vowels(s) from t2;
+-----------+--------------------+
| s         | udfs.has_vowels(s) |
+-----------+--------------------+
| lower     | true               |
| UPPER     | true               |
| Init cap  | true               |
| CamelCase | true               |
+-----------+--------------------+
Returned 4 row(s) in 0.24s

[localhost:21000] > create function count_vowels (string) returns int location '/user/hive/udfs/libudfsample.so' symbol='CountVowels';
[localhost:21000] > select count_vowels('cat in the hat');
+-------------------------------------+
| udfs.count_vowels('cat in the hat') |
+-------------------------------------+
| 4                                   |
+-------------------------------------+
Returned 1 row(s) in 0.12s
[localhost:21000] > select s, count_vowels(s) from t2;
+-----------+----------------------+
| s         | udfs.count_vowels(s) |
+-----------+----------------------+
| lower     | 2                    |
| UPPER     | 2                    |
| Init cap  | 3                    |
| CamelCase | 4                    |
+-----------+----------------------+
Returned 4 row(s) in 0.23s
[localhost:21000] > select count_vowels(null);
+-------------------------+
| udfs.count_vowels(null) |
+-------------------------+
| NULL                    |
+-------------------------+
Returned 1 row(s) in 0.12s

[localhost:21000] > create function strip_vowels (string) returns string location '/user/hive/udfs/libudfsample.so' symbol='StripVowels';
[localhost:21000] > select strip_vowels('abcdefg');
+------------------------------+
| udfs.strip_vowels('abcdefg') |
+------------------------------+
| bcdfg                        |
+------------------------------+
Returned 1 row(s) in 0.11s
[localhost:21000] > select strip_vowels('ABCDEFG');
+------------------------------+
| udfs.strip_vowels('abcdefg') |
+------------------------------+
| BCDFG                        |
+------------------------------+
Returned 1 row(s) in 0.12s
[localhost:21000] > select strip_vowels(null);
+-------------------------+
| udfs.strip_vowels(null) |
+-------------------------+
| NULL                    |
+-------------------------+
Returned 1 row(s) in 0.16s
[localhost:21000] > select s, strip_vowels(s) from t2;
+-----------+----------------------+
| s         | udfs.strip_vowels(s) |
+-----------+----------------------+
| lower     | lwr                  |
| UPPER     | PPR                  |
| Init cap  | nt cp                |
| CamelCase | CmlCs                |
+-----------+----------------------+
Returned 4 row(s) in 0.24s

Sample C++ UDA: SumOfSquares

This example demonstrates a user-defined aggregate function (UDA) that produces the sum of the squares of its input values.

The coding for a UDA is a little more involved than a scalar UDF, because the processing is split into several phases, each implemented by a different function. Each phase is relatively straightforward: the "update" and "merge" phases, where most of the work is done, read an input value and combine it with some accumulated intermediate value.

As in our sample UDF from the previous example, we add function signatures to a header file (in this case, uda-sample.h). Because this is a math-oriented UDA, we make two versions of each function, one accepting an integer value and the other accepting a floating-point value.

void SumOfSquaresInit(FunctionContext* context, BigIntVal* val);
void SumOfSquaresInit(FunctionContext* context, DoubleVal* val);

void SumOfSquaresUpdate(FunctionContext* context, const BigIntVal& input, BigIntVal* val);
void SumOfSquaresUpdate(FunctionContext* context, const DoubleVal& input, DoubleVal* val);

void SumOfSquaresMerge(FunctionContext* context, const BigIntVal& src, BigIntVal* dst);
void SumOfSquaresMerge(FunctionContext* context, const DoubleVal& src, DoubleVal* dst);

BigIntVal SumOfSquaresFinalize(FunctionContext* context, const BigIntVal& val);
DoubleVal SumOfSquaresFinalize(FunctionContext* context, const DoubleVal& val);

We add the function bodies to a C++ source file (in this case, uda-sample.cc):

void SumOfSquaresInit(FunctionContext* context, BigIntVal* val) {
  val->is_null = false;
  val->val = 0;
}
void SumOfSquaresInit(FunctionContext* context, DoubleVal* val) {
  val->is_null = false;
  val->val = 0.0;
}

void SumOfSquaresUpdate(FunctionContext* context, const BigIntVal& input, BigIntVal* val) {
  if (input.is_null) return;
  val->val += input.val * input.val;
}
void SumOfSquaresUpdate(FunctionContext* context, const DoubleVal& input, DoubleVal* val) {
  if (input.is_null) return;
  val->val += input.val * input.val;
}

void SumOfSquaresMerge(FunctionContext* context, const BigIntVal& src, BigIntVal* dst) {
  dst->val += src.val;
}
void SumOfSquaresMerge(FunctionContext* context, const DoubleVal& src, DoubleVal* dst) {
  dst->val += src.val;
}

BigIntVal SumOfSquaresFinalize(FunctionContext* context, const BigIntVal& val) {
  return val;
}
DoubleVal SumOfSquaresFinalize(FunctionContext* context, const DoubleVal& val) {
  return val;
}

As with the sample UDF, we build a shared library and put it into HDFS:

$ make
[  0%] Generating udf_samples/uda-sample.ll
[ 16%] Built target uda-sample-ir
Scanning dependencies of target udasample
[ 33%] Building CXX object CMakeFiles/udasample.dir/uda-sample.o
Linking CXX shared library udf_samples/libudasample.so
[ 33%] Built target udasample
Scanning dependencies of target uda-sample-test
[ 50%] Building CXX object CMakeFiles/uda-sample-test.dir/uda-sample-test.o
Linking CXX executable udf_samples/uda-sample-test
[ 50%] Built target uda-sample-test
[ 50%] Generating udf_samples/udf-sample.ll
[ 66%] Built target udf-sample-ir
[ 83%] Built target udfsample
[100%] Built target udf-sample-test
$ hdfs dfs -put ./udf_samples/libudasample.so /user/hive/udfs/libudasample.so

To create the SQL function, we issue a CREATE AGGREGATE FUNCTION statement and specify the underlying C++ function names for the different phases:

[localhost:21000] > use udf_testing;

[localhost:21000] > create table sos (x bigint, y double);
[localhost:21000] > insert into sos values (1, 1.1), (2, 2.2), (3, 3.3), (4, 4.4);
Inserted 4 rows in 1.10s

[localhost:21000] > create aggregate function sum_of_squares(bigint) returns bigint
  > location '/user/hive/udfs/libudasample.so'
  > init_fn='SumOfSquaresInit'
  > update_fn='SumOfSquaresUpdate'
  > merge_fn='SumOfSquaresMerge'
  > finalize_fn='SumOfSquaresFinalize';

[localhost:21000] > -- Compute the same value using literals or the UDA;
[localhost:21000] > select 1*1 + 2*2 + 3*3 + 4*4;
+-------------------------------+
| 1 * 1 + 2 * 2 + 3 * 3 + 4 * 4 |
+-------------------------------+
| 30                            |
+-------------------------------+
Returned 1 row(s) in 0.12s
[localhost:21000] > select sum_of_squares(x) from sos;
+------------------------+
| udfs.sum_of_squares(x) |
+------------------------+
| 30                     |
+------------------------+
Returned 1 row(s) in 0.35s

Until we create the overloaded version of the UDA, it can only handle a single data type. To allow it to handle DOUBLE as well as BIGINT, we issue another CREATE AGGREGATE FUNCTION statement:

[localhost:21000] > select sum_of_squares(y) from sos;
ERROR: AnalysisException: No matching function with signature: udfs.sum_of_squares(DOUBLE).

[localhost:21000] > create aggregate function sum_of_squares(double) returns double
  > location '/user/hive/udfs/libudasample.so'
  > init_fn='SumOfSquaresInit'
  > update_fn='SumOfSquaresUpdate'
  > merge_fn='SumOfSquaresMerge'
  > finalize_fn='SumOfSquaresFinalize';

[localhost:21000] > -- Compute the same value using literals or the UDA;
[localhost:21000] > select 1.1*1.1 + 2.2*2.2 + 3.3*3.3 + 4.4*4.4;
+-----------------------------------------------+
| 1.1 * 1.1 + 2.2 * 2.2 + 3.3 * 3.3 + 4.4 * 4.4 |
+-----------------------------------------------+
| 36.3                                          |
+-----------------------------------------------+
Returned 1 row(s) in 0.12s
[localhost:21000] > select sum_of_squares(y) from sos;
+------------------------+
| udfs.sum_of_squares(y) |
+------------------------+
| 36.3                   |
+------------------------+
Returned 1 row(s) in 0.35s

Typically, you use a UDA in queries with GROUP BY clauses, to produce a result set with a separate aggregate value for each combination of values from the GROUP BY clause. Let's change our sample table to use 0 to indicate rows containing even values, and 1 to flag rows containing odd values. Then the GROUP BY query can return two values, the sum of the squares for the even values, and the sum of the squares for the odd values:

[localhost:21000] > insert overwrite sos values (1, 1), (2, 0), (3, 1), (4, 0);
Inserted 4 rows in 1.24s

[localhost:21000] > -- Compute 1 squared + 3 squared, and 2 squared + 4 squared;
[localhost:21000] > select y, sum_of_squares(x) from sos group by y;
+---+------------------------+
| y | udfs.sum_of_squares(x) |
+---+------------------------+
| 1 | 10                     |
| 0 | 20                     |
+---+------------------------+
Returned 2 row(s) in 0.43s

Security Considerations for User-Defined Functions

When the Impala authorization feature is enabled:

  • To call a UDF in a query, you must have the required read privilege for any databases and tables used in the query.
  • Because incorrectly coded UDFs could cause performance or capacity problems, for example by going into infinite loops or allocating excessive amounts of memory, only an administrative user can create UDFs. That is, to execute the CREATE FUNCTION statement requires the ALL privilege on the server.

See Enabling Sentry Authorization for Impala for details about authorization in Impala.

Limitations and Restrictions for Impala UDFs

The following limitations and restrictions apply to Impala UDFs in the current release:

  • Impala does not support Hive UDFs that accept or return composite or nested types, or other types not available in Impala tables.
  • The Hive current_user() function cannot be called from a Java UDF through Impala.

  • All Impala UDFs must be deterministic, that is, produce the same output each time when passed the same argument values. For example, an Impala UDF must not call functions such as rand() to produce different values for each invocation. It must not retrieve data from external sources, such as from disk or over the network.
  • An Impala UDF must not spawn other threads or processes.
  • Prior to CDH 5.7 / Impala 2.5 when the catalogd process is restarted, all UDFs become undefined and must be reloaded. In CDH 5.7 / Impala 2.5 and higher, this limitation only applies to older Java UDFs. Re-create those UDFs using the new CREATE FUNCTION syntax for Java UDFs, which excludes the function signature, to remove the limitation entirely.
  • Impala currently does not support user-defined table functions (UDTFs).
  • The CHAR and VARCHAR types cannot be used as input arguments or return values for UDFs.

Converting Legacy UDFs During Upgrade to CDH 5.12 or Higher

In CDH 5.7 / Impala 2.5 and higher, new syntax is available for creating Java-based UDFs. UDFs created with the new syntax persist across Impala restarts, and are more compatible with Hive UDFs. Because the replication features in CDH 5.12 and higher only work with the new-style syntax, convert any older Java UDFs to use the new syntax at the same time you upgrade to CDH 5.12 or higher.

Follow these steps to convert old-style Java UDFs to the new persistent kind:

  • Use SHOW FUNCTIONS to identify all UDFs and UDAs.

  • For each function, use SHOW CREATE FUNCTION and save the statement in a script file.

  • For Java UDFs, change the output of SHOW CREATE FUNCTION to use the new CREATE FUNCTION syntax (without argument types), which makes the UDF persistent.

  • For each function, drop it and re-create it, using the new CREATE FUNCTION syntax for all Java UDFs.