1. In-Memory OLTP in SQL Server 2016 and Azure SQL Database
Microsoft Ignite 2016
2/4/2018 8:58 PM
BRK3097
In-Memory OLTP in SQL Server 2016 and Azure SQL Database
Jos de
Senior Program Manager
© 2016 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

2. Agenda Motivation and review of In-Memory OLTP
In-Memory OLTP in Azure SQL DB
Improvements in SQL Server 2016

3. There’s an opportunity to drive smarter decisions with data
The explosion of data sources...
25B
4.0B
1.3B
2010
2013
2020
There’s an opportunity to drive smarter decisions with data
…drives an explosion of data
CAGR = 41%
…which drives businesses to learn more and do more faster
Source: Forecast: Internet of Things, Endpoints and Associated Services, Worldwide, Gartner. Oct
Source: IDC “Digital Universe”, Dec. 2012

4. SQL In-Memory Technologies
2/4/2018
SQL In-Memory Technologies
SQL Real-Time Analytics
In 2016 and Azure DB
Faster Transactions
Faster Analytics +
IN-MEMORY OLTP +
IN-MEMORY DW
Up to 30x faster transaction processing with In-Memory OLTP
Over 100x query speed and significant data compression with
In-Memory ColumnStore
Using the same tables
© 2012 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

5. Why In-Memory OLTP
Customer demand for ever higher concurrency and lower latency, at a lower cost
Hardware trends demand architectural changes on RDBMS to meet those demands (large memory, many CPU cores)
In-memory OLTP is:
High performance,
Memory-optimized OLTP engine,
Integrated into SQL Server and
Architected for modern hardware

6. SQL Server In-Memory OLTP
Optimized for OLTP workloads
Memory-optimized data structures for efficient data access
Lock/Latch free access for linear scaling
Native Compilation of T-SQL Modules for efficient execution
Fully integrated into SQL Server
No new license or product
Familiar tools
Integrated Security and High Availability
No/Limited Application Changes
Available in SQL Server and Azure SQL DB
Performance numbers
Up to 30X performance improvement compared with traditional tables and stored procs
Actual gain depends on app pattern; typical gains: 2X-10X
Perf numbers in testing (4S server):
10M rows ingested per second
1.2GB/s of log generation
360K transactions per second for order processing workload
bwin.party experiences with ASP.NET session state
Consolidate 18 servers -> 1 in production
1M requests/sec in testing

7. In-Memory OLTP features
TechReady 23
2/4/2018 8:58 PM
In-Memory OLTP features
Memory-Optimized Tables
Data storage in memory-optimized structures
Compiled into DLLs for efficient data access
Default SCHEMA_AND_DATA (fully logged); support SCHEMA_ONLY (not logged)
Memory-Optimized Table Types
Memory-optimized table variables live in memory, in user DB (no tempdb, no IO)
Can be used for TVPs and table variables in all stored procs (not just native)
Natively Compiled Stored Procedures
Optimize performance of OLTP-style operations
Not suitable for reporting/analytics style queries
Natively Compiled Scalar User-Defined Functions
Do not require memory-optimized tables if there is no data access
© 2016 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

8. Customer Performance Gains
Data insertion at NoSQL speeds, with real-time analytics
Bwin.party – consolidate 18 SQL Servers -> 1;
1M requests/sec with 2016
EdgeNet – 10X perf gain for data ingestion
SBI Liquidity – 2X throughput; 4X latency improvement
CJ E&M – 35X transaction throughput

9. In-Memory OLTP Perf Gain Demo
Sources:
In-Memory OLTP perf demo
Memory-optimized table variables and temp tables

10. Memory-Optimized Indexes
Exist only in memory
Rebuilt on database startup (except for columnstore)
Do not contain data rows
Indexes contain memory pointers to the data rows
All indexes are covering (no bookmark lookups, no included columns)
Indexes are self-maintaining
Nonclustered Indexes
Dynamic index – grows/shrinks with data
No need to specify any options
Optimized for range and ordered scans
Hash Indexes
Static index – size does not change
Specify bucket_count – usually 1-2X the number of unique index keys
Optimized for point-lookups and inserts
Columnstore Indexes
Secondary copy of data, in compressed columnstore format
Optimized for analytics: large scans and aggregation – see BRK3071

11. Index Considerations Choosing HASH vs. NONCLUSTERED
TechReady 18
2/4/2018
Index Considerations
Choosing HASH vs. NONCLUSTERED
NONCLUSTERED is the safe baseline, as it supports all operations
Use HASH to optimize inserts and point lookups – inequality seeks and ordered scans not supported
You can have both types of indexes on the same columns
Hash Index Bucket Count
Choose at least as many buckets as unique index key values – over-sizing is usually OK
Automatically rounds up to the next power of 2
In case of many duplicates use a NONCLUSTERED index
Hash Index Key Columns
Does not support seek on subset of key columns
Index key must include exactly the columns you are searching on
Nonclustered Index Key: Choose direction (ASC/DESC) that conforms to ORDER BY requirement
© 2014 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

12. Accessing Memory Optimized Tables
TechReady 16
2/4/2018
Accessing Memory Optimized Tables
Natively Compiled Stored Procs
Access only memory optimized tables
Maximum performance for OLTP
Not suitable for analytics
Limitations on T-SQL surface area
When to use
OLTP-style operations
DML (insert/update/delete)
Queries touch limited number of rows
The more statements and the more complex the logic in a single proc -> the better the perf benefits
Interpreted T-SQL Access
Access both memory- and disk-based tables
Maximum performance for analytics
Virtually full T-SQL surface
When to use
Ad hoc queries
Reporting-style queries
Queries touching large nrs of rows (e.g., table scans)
Parallel queries (native modules are single-threaded)
Recommendation:
Use native compilation for most perf-critical stored procedures
Test performance of native vs. interpreted T-SQL
© 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries.
The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

13. In-Memory OLTP in Azure SQL Database

14. In-Memory OLTP for Azure SQL DB
In Public Preview for New Premium Databases
Planned: GA (H2 CY2016) to support Elastic Pools
Managed Service
No need to manage filegroup
No need to bind database to a resource pool for memory management
Data size:
P1 – 1GB user data size
P2 – 2GB user data size
P6 – 8GB user data size
P11 – 14GB user data size
P15 – 30GB user data size
Full In-Memory OLTP Surface Area
Includes all SQL Server 2016 features
New features are released to Azure SQL DB as soon as they are completed
CASE and sp_rename in the pipeline

15. Tooling Improvements

16. Performance improvements
Throughput:
Sustained log generation rate 1GB/s
Numerous improvements under the hood
Persisted Main Memory for Tail of the Log
Parallel plans and parallel scan of memory-optimized tables and indexes
Heap scan

17. Programmability improvements
Tables
Collations:
Full collations support in index key columns
All code pages supported for (var)char columns
Constraints
FOREIGN KEY
CHECK
UNIQUE constraints and indexes
LOB types: [n]varchar(max) and varbinary(max)
Row size >8060 bytes
NULLable index key columns
DML triggers on memory-optimized tables (AFTER)
Row-Level Security
Always Encrypted
Temporal Tables (history lives on-disk)
MARS (Multiple Active Result Sets) support
Native Compilation
Full collations support in native modules
Query surface area
{LEFT|RIGHT} OUTER JOIN
Disjunction (OR, NOT)
UNION [ALL]
SELECT DISTINCT
Subqueries (EXISTS, IN, scalar)
Nested Stored procedures (EXECUTE)
Natively Compiled Scalar UDFs
Access from both native and interop
Improves perf of scalar UDFs in traditional workload if no table access is required
Natively Compiled Inline TVFs
EXECUTE AS CALLER
Query Store

18. Manageability improvements
Improved SSMS experience
Lightweight Performance Analysis in SSMS
Generate migration checklists to evaluate all tables/procedures for migration
ALTER support for procedures and tables
Full ALTER TABLE/PROCEDURE support
Index management through ALTER TABLE … ADD/ALTER/DROP INDEX
Limitation: ALTER TABLE offline for now; ALTER of 10GB table takes 1.5 minutes in RTM (improvements inbound)
No ALTER between disk-based and memory-optimized
TDE (Transparent Data Encryption)
All on-disk data files are now encrypted once TDE is enabled
Stats improvements: auto-update and sampled stats
Statistics management is not on par with disk-based tables
No limits on data size
SQL2014 limitation: 256GB in durable tables
SQL2016 has no limitation, other than available storage (i.e., memory)
Windows Server 2016 supports up to 24TB of memory

19. New T-SQL surface area
Jos de Bruijn

20. Remaining Unsupported Features
Tables
DDL triggers and Transactional DDL
Data types: XML, Spatial, CLR types, datetimeoffset, rowversion, sql_variant
ALTER TABLE ONLINE
Cross-container transaction limitations: snapshot/snapshot, serializable/serializable; SAVEPOINT
Online Migration of disk-based tables to memory- optimized
Sp_rename
Interpreted T-SQL: cross-database queries, MERGE INTO, locking hints
Native Compilation
CASE, MERGE, JOIN in UPDATE/DELETE; VIEWs
Natively Compiled Table-Valued Functions
Automatic and statement-level recompile
Access to disk-based tables
Must be schema-bound; no dynamic T-SQL
Parallelism; limitations in query operators (hash join/agg, merge join)
Management: replication; DB snapshot; CDC; compression

21. Recap In-Memory OLTP offers extreme transaction processing performance
Memory-optimized data access and native compilation for low latency
Lock- and latch-free architecture for linear scaling with hardware resources
SQL Server 2016 offers big improvements
Much easier to use due to programmability and manageability improvements
Big enhancements in performance and scalability
Now in Azure SQL Database (Premium DB)
Now in Preview; general availability soon*

22. Related Content Where and how to use In-Memory OLTP
9/27 9am BRK3030: Review Columnstore Index in SQL Server 2016 and Azure SQL DB
9/27 10:45am BRK3097: Review In-Memory OLTP in SQL Server 2016 and Azure SQL DB
9/27 2:15pm BRK3071: Enable operational analytics in SQL Server 2016 and Azure SQL DB
9/28 2:00pm BRK2231: Understand how ChannelAdvisor is using SQL Server 2016
9/28 4pm BRK2007: Accelerate SQL Server 2016 HTAP performance with Windows 2016 and HPE
9/29 9am BRK3223: Understand how Attom Data Solutions is using SQL Server 2016 to accelerate their business
9/29 2:15pm BRK3103: Explore In-Memory OLTP architectures and customer case studies
9/30 10:45am BRK2083: Hear customer success stories with columnstore index in SQL Server 2016
9/30 12:30pm BRK3094: Accelerate SQL Server 2016 to the max: lessons learned from customer engagements
Where and how to use
In-Memory OLTP

23. Resources Samples Documentation Migrating apps Related videos
2/4/2018 8:58 PM
Resources
Samples
Perf demo
WideWorldImporters sample DB
IoT sample leveraging temporal tables (release page)
Azure DB Sample
Enable database for In-Memory OLTP
SQL Server Samples GitHub Repository
Documentation
Blog on In-Memory OLTP in SQL 2016
Memory-optimized table variables and temp tables
MSDN Documentation for In-Memory OLTP
Azure DB documentation for In-Memory
What’s new in SQL2016
Migrating apps
In-Memory OLTP Common Workloads and Migration Considerations
Migrating to In-Memory OLTP
Guidelines for using Indexes
Related videos
Microsoft Virtual Academy talks
Recent investments in In-Memory OLTP
© 2013 Microsoft Corporation. All rights reserved. Microsoft, Windows, and other product names are or may be registered trademarks and/or trademarks in the U.S. and/or other countries. The information herein is for informational purposes only and represents the current view of Microsoft Corporation as of the date of this presentation. Because Microsoft must respond to changing market conditions, it should not be interpreted to be a commitment on the part of Microsoft, and Microsoft cannot guarantee the accuracy of any information provided after the date of this presentation. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

24. Appendix: demo script use WideWorldImporters go -- verify indexes
-- new table functionality
DROP TABLE IF EXISTS dbo.Products GO
DROP TABLE IF EXISTS dbo.Orders
ProductID int IDENTITY PRIMARY KEY NONCLUSTERED, (
CREATE TABLE dbo.Products
-- character column with non-Latin code page
-- multiple large columns
ListPrice numeric(34,18) NULL,
Name varchar(100) COLLATE Frisian_100_CI_AI NOT NULL,
Annotations nvarchar(4000),
Description nvarchar(4000),
Notes nvarchar(max),
-- LOB columns
Comments nvarchar(4000),
CONSTRAINT CHK_ProductListPrice CHECK (ListPrice >= 0),
-- CHECK constraint
INDEX ix_Name (Name)
-- index on non-BIN2 column(s)
) WITH (MEMORY_OPTIMIZED=ON)
-- add index after table creation
ADD INDEX ix_ListPrice (ListPrice)
ALTER TABLE dbo.Products
-- index on nullable column(s)
CREATE TABLE dbo.Orders
Quantity int NOT NULL,
ProductID int NOT NULL INDEX ix_ProductID,
OrderID bigint IDENTITY,
NONCLUSTERED HASH (OrderID) WITH (BUCKET_COUNT= ),
CONSTRAINT PK_Orders PRIMARY KEY
-- foreign key
FOREIGN KEY (ProductID) REFERENCES dbo.Products(ProductID)
CONSTRAINT FK_Orders_ProductID
-- change hash index bucket count
REBUILD WITH (BUCKET_COUNT= )
ALTER INDEX PK_Orders
ALTER TABLE dbo.Orders
SELECT name, index_id, bucket_count
-- verify hash indexes
WHERE object_id=OBJECT_ID(N'dbo.Orders')
FROM sys.hash_indexes
CREATE FUNCTION dbo.ufn_GetSalesOrderStatusText tinyint)
-- natively compiled scalar user-defined function AS
WITH NATIVE_COMPILATION, SCHEMABINDING
RETURNS nvarchar(15)
BEGIN ATOMIC WITH (TRANSACTION ISOLATION LEVEL = SNAPSHOT, LANGUAGE = N'English')
RETURN 'Approved'
RETURN 'In Process'
RETURN 'Shipped'
RETURN 'Rejected'
RETURN 'Backordered'
RETURN 'Cancelled'
RETURN '** Invalid **'
END
-- =============================================================== auto-update stats
-- inspect plan
select * from dbo.Orders o join dbo.Products p on o.ProductID=p.ProductID
DBCC SHOW_STATISTICS ('dbo.Orders', ix_ProductID)
-- inspect stats
SET NOCOUNT ON
-- insert dummy data
INSERT dbo.Products (Name) VALUES ('Product1')
BEGIN TRAN
BEGIN
< 10000
int = 0
+= 1
INSERT dbo.Orders VALUES (1,1)
COMMIT

25. Free IT Pro resources To advance your career in cloud technology
Microsoft Ignite 2016
2/4/2018 8:58 PM
Free IT Pro resources To advance your career in cloud technology
Plan your career path
Microsoft IT Pro Career Center
Cloud role mapping
Expert advice on skills needed
Self-paced curriculum by cloud role
$300 Azure credits and extended trials
Pluralsight 3 month subscription (10 courses)
Phone support incident
Weekly short videos and insights from Microsoft’s leaders and engineers
Connect with community of peers and Microsoft experts
Get started with Azure
Microsoft IT Pro Cloud Essentials
Demos and how-to videos
Microsoft Mechanics
Connect with peers and experts
Microsoft Tech Community
© 2016 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

26. Please evaluate this session
2/4/2018 8:58 PM
Please evaluate this session
Your feedback is important to us!
From your PC or Tablet visit MyIgnite at
From your phone download and use the Ignite Mobile App by scanning the QR code above or visiting
© 2014 Microsoft Corporation. All rights reserved. MICROSOFT MAKES NO WARRANTIES, EXPRESS, IMPLIED OR STATUTORY, AS TO THE INFORMATION IN THIS PRESENTATION.

27. Thank you!