Tuesday Aug 25, 2015

Tips on SQL Plan Management and Oracle Database In-Memory - Part 2

In Part 1 of this series of tips on SQL Plan Management (SPM) and Oracle Database In-Memory, I covered what would happen if we have a SQL plan baseline for a full table scan query when the table was populating the In-Memory column store. 

In this part I’m going to cover a scenario where a query has more than one SQL plan baseline: 

  • There is a query (called Q2, for short).
  • Q2 queries a table called MYSALES, which is not yet populating the In-Memory column store.
  • Q2 filters rows in MYSALES using a predicate on the SALE_TYPE column.
  • Data in SALE_TYPE is skewed, so there’s an index and a histogram on this column.
  • Because there is data skew, Q2 has two accepted SQL plan baselines; one with a full table scan and one with an index range scan.

You’ve probably come across this situation many times: the Oracle Optimizer must choose between a full table scan or an index range scan depending on predicate selectivity. The ability to change the execution plan based on the value of bind variables is called adaptive cursor sharing. If you’ve not come across that, then you’ll find it useful to check out the section on this topic in the Database SQL Tuning Guide.

What’s great about SPM is that it allows you to have multiple SQL plan baselines for individual queries, so you're not forced to pick one plan in preference to another. This capability is most relevant in environments where SQL statements use bind variables and there is a good deal of data skew. Queries like this are likely to have their plans affected by Oracle In-Memory Database because in-memory full table scans will have a lower cost than storage-resident table scans. Clearly, the In-Memory column store will affect the point of inflection where a full table scan will become more efficient than an index range scan. How is this going to work with SPM? 

Take a look at the following example. Q2 executes and matches 2 million rows because I picked the value of bind variable “:val” to do just that. The Optimizer chooses a full table scan: 

PLAN_TABLE_OUTPUT
-----------------
SQL_ID  d3u63rk540w0r, child number 1
-------------------------------------
select /* SPM */ count(*),sum(val) from mysales where sale_type = :val

Plan hash value: 3292460164

------------------------------------------------------------------------------
  Id  | Operation          | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------
|   0 | SELECT STATEMENT   |         |       |       |  2475 (100)|          |
|   1 |  SORT AGGREGATE    |         |     1 |    17 |            |          |
|*  2 |   TABLE ACCESS FULL| MYSALES |  2000K|    32M|  2475   (1)| 00:00:01 |
------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------

  2 - filter("SALE_TYPE"=:VAL)

Note
-----
  - SQL plan baseline SQL_PLAN_93ct9zmnvtbuhc69cec1f used for this statement

For the second execution, the value “:val” is set so that it would match only 20,001 rows. This time the Optimizer chooses an index range scan: 

PLAN_TABLE_OUTPUT
-----------------
SQL_ID  d3u63rk540w0r, child number 2
-------------------------------------
select /* SPM */ count(*),sum(val) from mysales where sale_type = :val

Plan hash value: 1266559460

------------------------------------------------------------------------------------------------
  Id  | Operation                            | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
------------------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT                     |         |       |       |   133 (100)|          |
|   1 |  SORT AGGREGATE                      |         |     1 |    17 |            |          |
|   2 |   TABLE ACCESS BY INDEX ROWID BATCHED| MYSALES | 20001 |   332K|   133   (0)| 00:00:01 |
|*  3 |    INDEX RANGE SCAN                  | SI      | 20001 |       |    44   (0)| 00:00:01 |
------------------------------------------------------------------------------------------------

Predicate Information (identified by operation id):a
---------------------------------------------------
  3 - access("SALE_TYPE"=:VAL)

Note
-----
  - SQL plan baseline SQL_PLAN_93ct9zmnvtbuh5d8bf80c used for this statement

As you will have figured out, the Optimizer has calculated that the index is less efficient than a full table scan when Q2 matches a large number of rows (2 million in this case) so we have two viable SQL execution plans for this query. Before I ran the queries above, I accepted two SQL plan baselines for Q2. You can see in the “note” sections above that two different baselines are used (one ending in “80c” and one ending in “c1f”). They can be seen in the dba_sql_plan_baselines view: 

SELECT plan_name,sql_text,enabled, accepted 
FROM   dba_sql_plan_baselines
WHERE  sql_text LIKE '%SPM%';

PLAN_NAME                           SQL_TEXT                                ENA ACC
----------------------------------- ----------------------------------      --- ---
SQL_PLAN_93ct9zmnvtbuhc69cec1f      select /* SPM */ count(*),sum(val)      YES YES
                                    from mysales where sale_type = :val           
SQL_PLAN_93ct9zmnvtbuh5d8bf80c      select /* SPM */ count(*),sum(val)      YES YES
                                    from mysales where sale_type = :val

We’re good shape here. The Optimizer is adapting the query execution plan to take into account bind variable values and data skew. What’s more, SPM is working with us and not against us because it is not forcing Q2 to use a single SQL execution plan.

What happens if we populate MYSALES into the In-Memory column store? 

-- Mark MYSALES with the In-Memory attribute
ALTER TABLE mysales INMEMORY;

-- Access MYSALES to trigger population into In-Memory column store
SELECT count(*) FROM mysales;

If we execute Q2 to match 2 million rows, the Optimizer continues to choose a full table scan: 

PLAN_TABLE_OUTPUT
-----------------
SQL_ID  d3u63rk540w0r, child number 1
-------------------------------------
select /* SPM */ count(*),sum(val) from mysales where sale_type = :val

Plan hash value: 3292460164

---------------------------------------------------------------------------------------
  Id  | Operation                   | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
---------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT            |         |       |       |   115 (100)|          |
|   1 |  SORT AGGREGATE             |         |     1 |    17 |            |          |
|*  2 |   TABLE ACCESS INMEMORY FULL| MYSALES |  2000K|    32M|   115  (20)| 00:00:01 |
---------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------
  2 - inmemory("SALE_TYPE"=:VAL)
      filter("SALE_TYPE"=:VAL)

Note
-----
  - SQL plan baseline SQL_PLAN_93ct9zmnvtbuhc69cec1f used for this statement

The full table scan is now annotated with INMEMORY, so we know that some or all of the data for MYSALES is scanned via the In-Memory column store. The “note” section reports that the same baseline is being used as before (ending in “c1f”). This is good news, and it’s the scenario that was covered in Part 1 of this series.  

What if we executed the query to match 20,001 rows? You can probably guess what’s coming; the Optimizer judges that the In-Memory scan is more efficient than the index range scan: 

PLAN_TABLE_OUTPUT
-----------------
SQL_ID  d3u63rk540w0r, child number 2
-------------------------------------
select /* SPM */ count(*),sum(val) from mysales where sale_type = :val

Plan hash value: 3292460164

---------------------------------------------------------------------------------------
  Id  | Operation                   | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
---------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT            |         |       |       |   115 (100)|          |
|   1 |  SORT AGGREGATE             |         |     1 |    17 |            |          |
|*  2 |   TABLE ACCESS INMEMORY FULL| MYSALES | 20001 |   332K|   115  (20)| 00:00:01 |
---------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------
  2 - inmemory("SALE_TYPE"=:VAL)
      filter("SALE_TYPE"=:VAL)

Note
-----
  - SQL plan baseline SQL_PLAN_93ct9zmnvtbuhc69cec1f used for this statement

Since there is a SQL plan baseline that allows a full table scan to be used, Q2 can use this access method straight away and we get immediate benefit from scanning the In-Memory column store!

Hold on a minute! Wasn’t that just a little bit too convenient? I arranged it so that there was a handy full-table-scan SQL plan baseline ready and waiting for when I "flipped the switch" and started using the In-Memory column store. This example might seem a little contrived, but it is a real-world example and I chose it to illustrate how SPM works together with both Oracle In-Memory Database and adaptive cursor sharing (and if you want more, there's an earlier blog on how adaptive cursor sharing interacts with SPM).

If, instead, I had started out with a single baseline that specified an index range scan, then this is the plan that would have been used even after MYSALES populated the In-Memory column store (and we would not have had an INMEMORY FULL scan). That’s not a bad thing; it is exactly what plan stability means and it is how SPM is meant to work. In the example above I made use of a couple of SQL execution plans that were validated and accepted before I initiated the In-Memory column store. In the more general case, where the Optimizer identifies a brand new execution plan for use with the In-Memory column store, we might want to validate it before we allow the database to use it in our critical application. How can we do that? Happily, it's what SPM evolution was built for, and it goes all the way back to the initial scenario I mentioned in Part 1. I'll cover the details in Part 3 (coming soon). 

If you want to try out this example for yourself, the scripts are in GitHub.

Wednesday Aug 12, 2015

Tips on SQL Plan Management and Oracle Database In-Memory Part 1

If you follow Oracle’s In-Memory blog then you probably came across a post mentioning how you should use SQL Plan Management when you’re upgrading to Oracle Database In-Memory. Whether you have read that post or not, you might be wondering what will happen if you have some SQL plan baselines and you begin to populate the In-Memory column store with a bunch of tables as used by those baselines. That’s what this post is about. Well, in fact, I’m going to break the topic up into a few posts because (as ever!) there is a little bit of subtlety to cover. Luckily, this will make your life easier rather than more difficult because you can get immediate benefit from In-Memory even if you don’t evolve SQL plan baselines on day one.  

When I started to think about this post I thought that I would start with the first scenario that probably comes to mind if you’re familiar with SQL Plan Management (SPM): 
  • The Optimizer comes up with a new execution plan for a SQL statement because something has changed, and Oracle Database In-Memory would be a very good example of that! 
  • If there’s a SQL plan baseline for the statement, the database will use the baseline execution plan and capture the new plan.
  • Where appropriate, the new plan will be validated and accepted using SQL plan evolution. 

I will get to that, but first it’s better to start with a couple of more subtle points. With this information in our back pocket it will be easier to understand (and explain) the more traditional aspects of SQL plan evolution in the context of Oracle Database In-Memory. 

Here, I will cover the following example:
  • There is a table called MYSALES that’s not yet populated into the In-Memory column store. 
  • A query (called “Q1”) includes a full table scan of MYSALES. There is no index on the table that’s useful to Q1. 
  • Q1 has an active SQL plan baseline.
  • MYSALES is subsequently populated into the In-Memory column store.

Let’s take a look at Q1 and its SQL execution plan before populating MYSALES into the In-Memory column store (and I'll explain the significance of the highlighted text further down)...

SQL_ID  4ss4zbb813250, child number 0
-------------------------------------
SELECT /* SPM */ COUNT(*) FROM   mysales WHERE  val = 'X'

Plan hash value: 3292460164

------------------------------------------------------------------------------
  Id  | Operation          | Name    | Rows  | Bytes | Cost  %CPU | Time     |
------------------------------------------------------------------------------
|   0 | SELECT STATEMENT   |         |       |       |    69 (100)|          |
|   1 |  SORT AGGREGATE    |         |     1 |     2 |            |          |
|*  2 |   TABLE ACCESS FULL| MYSALES | 99991 |   195K|    69   (2)| 00:00:01 |
------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------
  2 - filter("VAL"='X')

Note
-----
  - SQL plan baseline SQL_PLAN_7469nmnn7nsu3c69cec1f used for this statement

Q1 performs a full table scan of MYSALES. The "note" section makes it clear that a SQL plan baseline is used. This is what that looks like:

SELECT PLAN_TABLE_OUTPUT
FROM   V$SQL s, DBA_SQL_PLAN_BASELINES b,
      TABLE(
        DBMS_XPLAN.DISPLAY_SQL_PLAN_BASELINE(b.sql_handle,b.plan_name,'basic')
      ) t
WHERE  s.EXACT_MATCHING_SIGNATURE=b.SIGNATURE
AND    b.PLAN_NAME=s.SQL_PLAN_BASELINE
AND    s.SQL_ID='4ss4zbb813250';

--------------------------------------------------------------------------------
SQL handle: SQL_7219349d287a6343
SQL text: SELECT /* SPM */ COUNT(*) FROM   mysales WHERE  val = 'X'
--------------------------------------------------------------------------------

--------------------------------------------------------------------------------
Plan name: SQL_PLAN_7469nmnn7nsu3c69cec1f         Plan id: 3332172831
Enabled: YES     Fixed: NO      Accepted: YES     Origin: AUTO-CAPTURE
Plan rows: From dictionary
--------------------------------------------------------------------------------

Plan hash value: 3292460164

--------------------------------------
  Id  | Operation          | Name    |
--------------------------------------
|   0 | SELECT STATEMENT   |         |
|   1 |  SORT AGGREGATE    |         |
|   2 |   TABLE ACCESS FULL| MYSALES |
--------------------------------------

What happens if MYSALES is now populated into the In-Memory column store? 

-- Mark MYSALES with the In-Memory attribute
ALTER TABLE mysales INMEMORY;

-- Access MYSALES to trigger population into In-Memory column store
SELECT count(*) FROM mysales;

Let’s rerun our query and examine the execution plan:

SQL_ID  4ss4zbb813250, child number 1
-------------------------------------
SELECT /* SPM */ COUNT(*) FROM   mysales WHERE  val = 'X'

Plan hash value: 3292460164

---------------------------------------------------------------------------------------
  Id  | Operation                   | Name    | Rows  | Bytes | Cost (%CPU)| Time     |
---------------------------------------------------------------------------------------
|   0 | SELECT STATEMENT            |         |       |       |     3 (100)|          |
|   1 |  SORT AGGREGATE             |         |     1 |     2 |            |          |
|*  2 |   TABLE ACCESS INMEMORY FULL| MYSALES |   100K|   195K|     3   (0)| 00:00:01 |
---------------------------------------------------------------------------------------

Predicate Information (identified by operation id):
---------------------------------------------------
   2 - inmemory("VAL"='X')
       filter("VAL"='X')

Note
-----
   - SQL plan baseline SQL_PLAN_7469nmnn7nsu3c69cec1f used for this statement

There is still a full table scan, but this time the query will read data from MYSALES via the In-Memory column store rather than the storage-resident table and, even better, the same SQL plan baseline is used. That was pretty easy! The Optimizer chose a full table scan in both cases, so the same SQL plan baseline was used both cases. The INMEMORY annotation for the full table scan is “for your information only”; it tells you that the query scanned some or all of the data for your table via the In-Memory column store but as far as the Optimizer is concerned it is “just” a full table scan, as the keyword INMEMORY does not affect the plan hash value, so it will match the existing the SQL plan baseline (above, you can see that the plan hash value is always "3292460164" ).

Why do I say the INMEMORY keyword indicates some or all of the data for your table is scanned via the In-Memory column store? Remember until all of the data belonging to MYSALES has been populated into the In-Memory column store, Oracle will automatically pick up the rest of the data from wherever it resides. That could be from memory (e.g. the buffer cache) or from flash or from disk.

It should be pretty obvious by now that if we decide to remove MYSALES from the In-Memory column store, the query will revert to scanning the storage-resident table and the plan will display “TABLE ACCESS FULL”. 

This example is very simple, but the principle applies to queries that have the same execution plan for In-Memory versus non-In-Memory. What happens if there are execution plan changes and, in particular, if indexes are involved? Start by looking at Part 2.

If you want to try out this example for yourself, the scripts are in GitHub.


Wednesday Jun 24, 2015

What you need to know about SQL Plan Management and Auto Capture

SQL Plan Management (SPM) is an Oracle database feature that allows you to establish a set of SQL execution plans that will be used even if the database is subject to changes that would otherwise cause execution plan changes to occur. For example, you might have an end-of-day batch run that operates in a business context where there are extreme peaks and troughs in daily volume, or perhaps you are upgrading a database and want to be sure that plans are carried over (at least initially). You do not have to fix execution plans in stone with SPM, you can use plan evolution to automate the process of finding improved plans, improving performance in a controlled way and at your own pace. If you’re not familiar with SPM, a very good place to start is to take a look at Maria Colgan’s four-part blog post on the subject. It gives you all the tools you need to get started.

If you are using SPM in Oracle Database 11gR2 or 12c, or if you are considering whether you should use it, then this blog post is for you. I decided to publish this post because I recently encountered a couple of environments that ran into, let’s say, “difficulties” with SPM when capturing SQL plan baselines automatically and continuously over a very long period of time (more than a year in fact). I’d like to give you a few pointers to avoid running into the same problems and why automatic SQL baseline capture was never intended to be used in that way.

[Read More]

Tuesday May 26, 2015

Space Management and Oracle Direct Path Load

Most of you will be familiar with the concept of direct path load and how it’s an efficient way to load large volumes of data into an Oracle database as well as being a great technique to use when moving and transforming data inside the database. It’s easy to use in conjunction with parallel execution too, so you can scale out and use multiple CPUs and database servers if you need to process more data in less time.

Probably less well known is how the Oracle database manages space during direct path load operations. This is understandable because the Oracle database uses a variety of approaches and it has not always been very obvious which one it has chosen. The good news is that Oracle Database 12c from version 12.1.0.2 onwards makes this information visible in the SQL execution plan and it is also possible to understand what’s happening inside earlier releases with a bit of help from this post and some scripts I’ve linked to at the end.

[Read More]

Friday May 08, 2015

Controlling Access To The In-Memory Column Store Via Hints

It’s been almost a year since I’ve had an opportunity to write a new blog post here due to my crazy schedule now that I'm the In-Memory Maven (thanks to Doug Burns & Graham Wood for the new and improved title). But this morning while I was writing a post for the In-Memory blog about controlling the use of the In-Memory column store (IM column store), I realized that the content on Optimizer hints really belonged over here.

So, I decided to split the blog post in two. I’ve put the information on the initialization parameter that control the use of the IM column store on the In-Memory blog and the information about the Optimizer hint that control the use of the IM column store here.

[Read More]

Monday Jul 07, 2014

The Optimizer & the new Oracle Database In-Memory option

I know its been forever since I've posted a new blog but I am no longer the Optimizer lady. In case you haven't heard already,  I have a new role now as the In-Memory lady (definitely need to work on a better title).

But for this weeks In-Memory blog post I got a chance to combine both my old and my new roles when I wrote about how the In-Memory option and the Optimizer interact. And I thought the readers of this blog might be interested in this topic too.

So, if you are interested in seeing how these these two fascinating pieces of the Oracle technology work together, check out the latest post on the new In-Memory blog.

You should also mark your calendars because of the head of the Optimizer development team, Mohamed Zait, will also discuss this topic at Oracle Open World later this year!

 Hope to catch up with a lot of you then.

Friday Sep 27, 2013

Oracle Open World 2013 - It's a wrap

Thanks to all those that attended the Optimizer sessions and stopped by the Optimizer demopod during this years Oracle OpenWorld. It was an amazing conference this year with lots of great technical sessions and interesting discussion at the demo grounds.

Since so many folks have emailed me requesting a copy of the Optimizer session presentations I thought it would be a good idea to post them here. The presentation were unfortunately were larger than the 2MB file limit so I have broken them multiple files. Enjoy!

Session CON8622 :

Session CON8643 :


Wednesday Sep 25, 2013

Day 4 of Oracle OpenWorld 2013 September 25th

Thanks to all those who stopped by the demogrounds  to chat with the Optimizer developers and to check out what is new in the Oracle Optimizer over the last two days. Remember, today is the last day of the demogrounds, so if you haven't had a chance to stop by yet, do so today. The Optimizer developers will be there from 9:45 am until 4pm.

There are also some great technical session on today, including:

  • Oracle Database In-Memory—The Next Big Thing
    Session GEN9312 Presented by Juan Loaiza at 11:45am in Moscone South - room 103
    This session provides a deep-dive explanation of how the new Oracle Database In-Memory works. It's a must see for everyone interested in performance.

  • Optimizer Statistics: A Fresh Approach
    Session CON7442 Presented by Conor McDonald at Moscone South - room 308

Tuesday Sep 24, 2013

Day 3 of Oracle OpenWorld 2013 September 24th

Hopefully you enjoyed yesterday, the first full day of technical sessions at Oracle OpenWorld and are ready for more today!

Today we give our second technical session, Oracle Optimizer Boot Camp: 10 Optimizer Tips You Can’t Do Without (Session CON8643) at 5:15pm, in Moscone South - room 104.


In this session Jonathan Lewis will join us to explain the process of analyzing and solving 10 of the most common SQL execution performance problems. A must see for all those interested in the Optimizer and SQL tuning or England versus Ireland match ups!

The Optimizer team will also be at the Oracle Database Demogrounds all day.  Demogrounds open at 9:45 am and run until 6pm. So stop by and find out what's new with the Optimizer and the statistics that feed it.

Monday Sep 23, 2013

Day 2 of Oracle OpenWorld 2013 September 23rd

Oracle OpenWorld started yesterday and San Francisco is just buzzing with Oracle folks and the excitement generated by the new Oracle Database In-Memory announcements made by Larry Ellison during last nights keynote!

If you are attending the conference don't miss the opportunity to chat with the Optimizer development team at one of our technical sessions or at the Oracle Demo grounds.

Our first technical session, Oracle Optimizer: What’s New in Oracle Database 12c? Session CON8622 is on today at 3:15pm in Moscone North, room 131.

This session provides a deep-dive explanation of how the new adaptive approach to query optimization works. We will use a real-world demo to show the different components of the adaptive model and how they help the optimizer learn additional information during query execution that can be used to improve not only the performance of the current SQL statement but all statements.

Members of the Optimizer team will also be available from 9:30am to 6pm at the demogrounds, to answer any Optimizer questions you might have. This may also be your last change to snag the limited edition Optimizer bumper stickers!

+Maria Colgan

Sunday Sep 22, 2013

Day 1 of Oracle OpenWorld 2013 September 22nd

The biggest day of the Oracle calendar has arrived. Oracle OpenWorld 2013 officially starts today!

The show kicks off with ton's of great technical sessions selected by the Oracle User Groups including several sessions on the optimizer:

  • 11:45 AM - 12:45 PM The Query Optimizer in Oracle Database 12c: What’s New?
    Session UGF3062 delivered by Christian Antognini in Moscone West - room 3002

  • 1:00 PM - 2:00 PM Solving Critical Customer Issues with the Oracle Database 12c Optimizer
    Session UGF5498 delivered by Nicolas Jardot in Moscone West - room 3002

And of course, Larry's keynote is this evening 5:15pm – 7:00pm, Moscone North. A must see, as he is bound to make some exciting announcements to get the show started! One of which may be very close to my heart but you will have show up to find out what it is!

All right I will give you a hint .........


Hope to see ya there!

Friday Sep 13, 2013

What's new in 12c: Adaptive joins part 2

In our earlier post on adaptive joins we explained how this new 12c functionality works and said we would follow up this post with a real-world demo.

[Read More]

Sunday Sep 08, 2013

Oracle OpenWorld 2013 - Time to plan your schedule

There are only two weeks to go until Oracle Open World, the largest gathering of Oracle customers, partners, developers, and technology enthusiasts, which begins on September 22nd in San Francisco. Of course the Optimizer development group will be there and you will have multiple opportunities to meet the team, in one of our technical sessions, or at the Oracle demogrounds.

This year the Optimizer team has 2 technical sessions;

  • Monday, September 23rd at 3:15pm  Oracle Optimizer: What’s New in Oracle Database 12c?
    Session CON8622
    at Moscone North - room 131
    This session provides a deep-dive explanation of how the new adaptive approach to query optimization works. We will use a real-world demo to show the different components of the adaptive model and how they help the optimizer learn additional information during query execution that can be used to improve not only the performance of the current SQL statement but all statements.

  • Tuesday, September 24th at 5:15pm Oracle Optimizer Bootcamp: 10 Optimizer tips you can't do without
    Session CON8643 at Moscone South - room 104
    Jonathan Lewis will join us to explain the process of analyzing and solving 10 of the most common SQL execution performance problems. These problems include poor cardinality estimates, bind peeking issues, the selection of sub-optimal access methods, and many more. Through clear how-to examples, you will learn how to identify and quickly resolve these issues and add 10 new tricks to your SQL tuning arsenal.

If you have burning Optimizer or statistics related questions, you can ask them at the Optimizer demo booth in the Database area of the demogrounds. Members of the Optimizer development team will be there Monday to Wednesday from 9:45am until 5:30pm. You may even be able to pick up an optimizer bumper sticker.

The full searchable OOW catalog is on-line, or you can browse the speakers by name. So start planning your trip today!

Sunday Aug 25, 2013

What's new in 12c: Adaptive joins

As we promised in our previous post, we are starting a blog series describing all of new Optimizer and statistics related functionality on Oracle Database 12c. We begin the series with an in-depth look at adaptive plans, one of the key features in the new adaptive query optimization framework.[Read More]

Tuesday Jun 25, 2013

Oracle Database 12c is here!

Oracle Database 12c was officially release today and is now available for download. Along with the software release comes a whole new set of collateral that explains in detail all of the new features and functionality you will find in this release.

The Optimizer page on Oracle.com has all the juicy details about what you can expect from the Optimizer in Oracle Database 12c. Direct links are below.

[Read More]
About

The Oracle Optimizer blog is written by members of the Optimizer development team. The goal of this blog is to provide an insight into the workings of the Optimizer and the statistics it relies on. The views expressed on this blog are our own and do not necessarily reflect the views of Oracle and its affiliates. The views and opinions expressed by visitors on this blog are theirs solely and may not reflect ours.

Search

Archives
« August 2015
SunMonTueWedThuFriSat
      
1
2
3
4
5
6
7
8
9
10
11
13
14
15
16
17
18
19
20
21
22
23
24
26
27
28
29
30
31
     
Today