[DRILL-6238] Batch sizing for operators

• Type: New Feature
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Assignee: Padma Penumarthy
• Reporter: Padma Penumarthy
• Created: 13/Mar/18 23:14
• Updated: 09/Jul/18 17:34

Description

Batch Sizing For Operators

This document describes the approach we are taking for limiting batch sizes for operators other than scan.

Motivation

Main goals are

1. Improve concurrency - By having each query use less memory i.e. stay within budget, we can have more queries run concurrently. 
2. Reduce query failures because of out of memory errors

To accomplish these goals, we need to make queries execute within a specified memory budget. To enforce per query memory limit, we need to be able to enforce per fragment and per operator memory limits. Controlling individual operators batch sizes is the first step towards all this.

Background

In Drill, different operators have different limits w.r.to outgoing batches. Some use hard coded row counts, some use hard coded memory and some have none at all. Based on input data size and what the operator is doing, memory used by the outgoing batch can vary widely as there are no limits imposed. Queries fail because we are not able to allocate the memory needed. Some operators produce very large batches, causing blocking operators like sort, hash agg which have to work under tight memory constraints to fail. Size of batches should be a function of available memory rather than input data size and/or what the operator does. Please refer to table at the end of this document for details on what each operator does today.

Design

Goal is to have all operators behave the same way i.e. produce batches with size less than or equal to configured outgoing batch size with a minimum of 1 row per batch and maximum of 64k rows per batch. A new system option ‘drill.exec.memory.operator.output_batch_size’ is added which has default value of 16MB.

The basic idea is to limit size of outgoing batch by deciding how many rows we can have in the batch based on average entry size of each outgoing column, taking into account actual data size and metadata vector overhead we add on top for tracking variable length, mode(repeated, optional, required) etc. The calculations to figure out number of rows will be different for each operator and is based on

1. What the operator is doing
2. Incoming batch size that includes information on type and average size of each column
3. What is being projected out

By taking this adaptive approach based on actual average data sizes, for operators which were limiting batch size to less than 64K rows before can possibly do lot more rows (upto 64K rows) in a batch if the memory stays within the budget. For example, flatten and joins have batch size of 4K rows, which probably might have been done to be conservative w.r.to memory usage. By making these operators go upto 64K rows as long as they stay with in the memory budget should help improve performance.

Also, to improve performance and utilize memory more efficiently, we will

1. Allocate memory for value vectors upfront. Since we know the number of rows and sizing information for each column in the  outgoing batch, we will use that information to allocate memory for value vectors upfront.  Currently, we either do initial allocation for 4K values and keep doubling every time we need more or allocate for maximum needed upfront. With this change to pre allocate memory based on sizing calculation, we can improve performance by reducing the memory copies and zeroing the new half we do every time we double and help save memory in cases where we were over allocating before.
2. Round down the number of rows in outgoing batch to a power of two. Since memory is allocated in powers of two, this will help us pack the value vectors densely thereby reducing the amount of memory that gets wasted because of doubling effect.

So, to summarize, the benefits we will get are improved memory utilization, better performance, higher concurrency and less queries dying because of out of memory errors.

Note: Since these sizing calculations are based on averages, strict memory usage enforcement is not possible. There could be pathological cases where because of uneven data distribution, we might exceed the configured output batch size potentially causing OOM errors and problems in downstream operators.

Other issues that will be addressed:

• We are adding extra processing for each batch in each operator to figure out the sizing information. This overhead can be reduced by passing this information along with the batch between operators.
• For some operators, it will be complex to figure out average size of outgoing columns especially if we have to evaluate complex expression trees and UDFs to figure out the transformation on incoming batches. We will use approximations as appropriate.

Following table summarizes the limits we have today for each operator.

flatten, merge join and external sort adhere to batch size limits as described in this document as of drill release 1.13.

 

Operator	Limit (Rows, Memory)	Notes
Flatten	4K,  512MB	Flatten can produce very large batches based on average cardinality of the flatten column. Batch size limit in 1.13.
Merge Receiver	32K	No memory limit.
Hash Aggregate	64K	No memory limit.
Streaming Aggregate	32K	No memory limit. Streaming Aggregate typically does not produce large batches.
Broadcast Sender	None	No limits.
Filter, Limit	None	No limits. These operators just add selection vector (max of 256KB) on top of incoming batch. If incoming batch is with in batch size limit, it is not a priority to adjust outgoing batch size for these operators.
Hash Join	4K	No memory limit. Joins produce large batches. 4k limit might have been put in place to be conservative w.r.to memory usage.
Merge Join	4K	batch size limit is added in 1.13
Nested Loop Join	4K	No memory limit
Partition Sender	1K
Project	64K	No memory limit. Project can produce large batches if we are projecting lot of columns i.e. more than incoming. Also, some functions like concat can increase average row width of columns, using up lot more memory. It is also complex to figure out average row width of the columns in the outgoing batch because of complex expressions and UDFs.
Selection Vector Remover	None	No limits. Output batches are smaller than input batches. It is not a priority to limit output batch size for this operator.
TopN	4K	No memory limit. TopN typically does not produce large batches.
Union	None	No limit
Windows	None	No limit
External Sort	64k, 16 MB	External sort already limits the size of output batches to 16MB.
Unordered Receiver	None	No limit

Issue Links

• Dependent: DRILL-6232 Vector initializer used for memory allocation in external sort is subject to aliasing
• Dependent: DRILL-6411 Make batch memory sizing logs uniform across all operators
• Dependent: DRILL-6478 enhance debug logs for batch sizing
• Dependent: DRILL-6310 limit batch size for hash aggregate
• Dependent: DRILL-6356 batch sizing for union all
• Dependent: DRILL-6549 batch sizing for nested loop join
• is related to: DRILL-6160 Limit batch size for streaming aggregate based on memory
• is related to: DRILL-6113 Limit batch size for Merge Receiver
• is related to: DRILL-6184 Add batch sizing information to query profile
• is related to: DRILL-6161 Allocate memory for outgoing vectors based on sizing calculations
• is related to: DRILL-6180 Use System Option "output_batch_size" for External Sort
• is related to: DRILL-6071 Limit batch size for flatten operator
• is related to: DRILL-6123 Limit batch size for Merge Join based on memory
• is related to: DRILL-6162 Enhance record batch sizer to retain nesting information for map columns.
• is related to: DRILL-6177 Merge Join - Allocate memory for outgoing value vectors based on sizes of incoming batches.
• is related to: DRILL-6236 batch sizing for hash join