1. Understanding the value and
architecture of Apache Drill
Michael Hausenblas, Chief Data Engineer EMEA, MapR
Hadoop Summit, Amsterdam, 2013-03-20
1

2. Kudos to http://cmx.io/
2
2

3. Workloads
• Batch processing (MapReduce)
• Light-weight OLTP (HBase, Cassandra, etc.)
• Stream processing (Storm, S4)
• Search (Solr, Elasticsearch)
• Interactive, ad-hoc query and analysis (?)
3

4. Interactive Query at Scale
Impala
low-latency
4

5. Use Case
• Jane, a marketing analyst
• Determine target segments
• Data from different sources
5

6. Today’s Solutions
• RDBMS-focused
– ETL data from MongoDB and Hadoop
– Query data using SQL
• MapReduce-focused
– ETL from RDBMS and MongoDB
– Use Hive, etc.
6

7. Requirements
• Support for different data sources
• Support for different query interfaces
• Low-latency/real-time
• Ad-hoc queries
• Scalable and fast
• Reliable
7

8. Google’s Dremel
http://research.google.com/pubs/pub36632.html
8

9. Apache Drill Overview
• Inspired by Google’s Dremel
• Standard SQL 2003 support
• Other QL possible
• Plug-able data sources
• Support for nested data
• Schema is optional
• Community driven, open, 100’s involved
9

10. Apache Drill Overview
10

11. High-level Architecture
11

12. High-level Architecture
• Each node: Drillbit - maximize data locality
• Co-ordination, query planning, execution, etc, are distributed
• By default Drillbits hold all roles
• Any node can act as endpoint for a query
Drillbit Drillbit Drillbit Drillbit
Storage Storage Storage Storage
Process Process Process Process
node node node node
12

13. High-level Architecture
• Zookeeper for ephemeral cluster membership info
• Distributed cache (Hazelcast) for metadata, locality
information, etc.
Zookeeper
Drillbit Drillbit Drillbit Drillbit
Distributed Cache Distributed Cache Distributed Cache Distributed Cache
Storage Storage Storage Storage
Process Process Process Process
node node node node
13

14. High-level Architecture
• Originating Drillbit acts as foreman, manages query execution,
scheduling, locality information, etc.
• Streaming data communication avoiding SerDe
Zookeeper
Drillbit Drillbit Drillbit Drillbit
Distributed Cache Distributed Cache Distributed Cache Distributed Cache
Storage Storage Storage Storage
Process Process Process Process
node node node node
14

15. Principled Query Execution
Source Logical Physical
Query Parser Plan Optimizer Plan Execution
SQL 2003 parser API query: [
{
topology scanner API
DrQL @id: "log",
op: "sequence",
MongoQL do: [
{
DSL op: "scan",
source: “logs”
},
{
op:
"filter",
condition:
"x > 3”
},
15

16. Drillbit Modules
RPC Endpoint
SQL
Scheduler
Storage Engine Interface
DFS Engine
Physical Plan
Logical Plan
HiveQL
Optimizer Foreman
Pig HBase Engine
Operators
Mongo
Parser
Distributed Cache
16

17. Key Features
• Full SQL 2003
• Nested data
• Optional schema
• Extensibility points
17

18. Full SQL – ANSI SQL 2003
• SQL-like is often not enough
• Integration with existing tools
– Datameer, Tableau, Excel, SAP Crystal Reports
– Use standard ODBC/JDBC driver
18

19. Nested Data
• Nested data becoming prevalent
– JSON/BSON, XML, ProtoBuf, Avro
– Some data sources support it natively
(MongoDB, etc.)
• Flattening nested data is error-prone
• Extension to ANSI SQL 2003
19

20. Optional Schema
• Many data sources don’t have rigid schemas
– Schema changes rapidly
– Different schema per record (e.g. HBase)
• Supports queries against unknown schema
• User can define schema or via discovery
20

21. Extensibility Points
• Source query – parser API
• Custom operators, UDF – logical plan
• Optimizer
• Data sources and formats – scanner API
Source Logical Physical
Query Parser Plan Optimizer Plan Execution
21

22. … and Hadoop?
• HDFS can be a data source
• Complementary use cases …
• … use Apache Drill
– Find record with specified condition
– Aggregation under dynamic conditions
• … use MapReduce
– Data mining with multiple iterations
– ETL
https://cloud.google.com/files/BigQueryTechnicalWP.pdf
22
22

23. Example
{
"id": "0001",
"type": "donut",
”ppu": 0.55,
"batters":
{
{
"batter”:
"sales" : 700.0,
[
"typeCount" : 1,
{ "id": "1001", "type": "Regular" },
"quantity" : 700,
{ "id": "1002", "type": "Chocolate" },
"ppu" : 1.0
…
}
{
"sales" : 109.71,
data source: donuts.json "typeCount" : 2,
"quantity" : 159,
query:[ { "ppu" : 0.69
op:"sequence", }
do:[ {
{ "sales" : 184.25,
op: "scan", "typeCount" : 2,
ref: "donuts", "quantity" : 335,
source: "local-logs", "ppu" : 0.55
selection: {data: "activity"} }
},
{ result: out.json
op: "filter",
expr: "donuts.ppu < 2.00"
},
…
logical plan: simple_plan.json https://cwiki.apache.org/confluence/display/DRILL/Demo+HowTo
23

24. Status
• Heavy development by multiple organizations
• Available
– Logical plan (ADSP)
– Reference interpreter
– Basic SQL parser
– Basic demo
24

25. Status
March/April
• Larger SQL syntax
• Physical plan
• In-memory compressed data interfaces
• Distributed execution focused on large cluster
high performance sort, aggregation and join
• Storage engine implementations (HBase, etc.)
25

26. Contributing
• Dremel-inspired columnar format: Twitter’s Parquet and
Hive’s ORC file
• Integration with Hive metastore (?)
• DRILL-13 Storage Engine: Define Java Interface
• DRILL-15 Build HBase storage engine implementation
26

27. Contributing
• DRILL-48 RPC interface for query submission and physical plan
execution
• DRILL-53 Setup cluster configuration and membership mgmt
system
– ZK for coordination
– Helix for partition and resource assignment (?)
• Further schedule
– Alpha Q2
– Beta Q3
27

28. Kudos to …
• Julian Hyde, Pentaho
• Timothy Chen, Microsoft
• Chris Merrick, RJMetrics
• David Alves, UT Austin
• Sree Vaadi, SSS/NGData
• Jacques Nadeau, MapR
• Ted Dunning, MapR
28

29. Engage!
• Follow @ApacheDrill on Twitter
• Sign up at mailing lists (user|dev)
http://incubator.apache.org/drill/mailing-lists.html
• Learn where and how to contribute
https://cwiki.apache.org/confluence/display/DRILL/Contributing
• Keep an eye on http://drill-user.org/
29