RE: application monitoring best practices (sleeper features)

From: Iggy Fernandez <iggy_fernandez_at_hotmail.com>
Date: Wed, 28 Aug 2013 08:49:59 -0700
Message-ID: <BLU179-W373C975DC7922D49D3901CEB4B0_at_phx.gbl>



"Sleeper Features" would be an awesome article for the NoCOUG Journal if you would write one for us.

My pet "sleeper feature" is the oldest in the book: multi-table clusters in indexed and hash flavors. The impression I get is that even Oracle product management does not believe in it. Partitioned hash clusters were first used by Oracle in TPC-C benchmarks more than two years but have not been implemented in Oracle Database 12c. See http://iggyfernandez.wordpress.com/2011/05/10/major-new-undocumented-partitioning-feature-in-oracle-database-11g-release-2/. I believe that clusters hold the answers to many performance problems as well as the antidote to NoSQL. See http://iggyfernandez.wordpress.com/2013/07/28/no-to-sql-and-no-to-nosql/.

How about MDC (Multidimensional clustering)? Clusters again! See http://iggyfernandez.wordpress.com/2011/03/26/multidimensional-clustering-mdc-in-oracle-database/. Another sleeper feature!

Iggy

-- 
Iggy Fernandez
Email: iggy_fernandez_at_hotmail.com
Cellphone: (925) 478 3161
Blog: So Many Manuals So Little Time
Author of Beginning Oracle Database 11g Administration Editor of the NoCOUG Journal

> From: kevin.jernigan_at_oracle.com
>
> Don't get me started - it's part of my job to spread the word about the
> "sleeper" features in the database, such as temporal, CQN, in-db
> archiving, etc...KJ
>
> --
> Kevin Jernigan
> Senior Director Product Management
> Advanced Compression, Hybrid Columnar
> Compression (HCC), Database File System
> (DBFS), SecureFiles, Database Smart Flash
> Cache, Total Recall, Database Resource
> Manager (DBRM), Direct NFS Client (dNFS),
> Continuous Query Notification (CQN),
> Index Organized Tables (IOT), Information
> Lifecycle Management (ILM)
> (650) 607-0392 (o)
> (415) 710-8828 (m)
>
> On 8/27/13 1:14 PM, Mark Bobak wrote:

>> Damn, never even *heard* of that feature! Never ceases to amaze me how
>> many new Oracle features go unnoticed for who knows how long!
>>
>>
>> Thanks Kevin!
>>
>> -Mark
>> On 8/27/13 3:58 PM, "Kevin Jernigan" <kevin.jernigan_at_oracle.com> wrote:
>>
>>> Have you looked at Continuous Query Notification - CQN
>>> <http://docs.oracle.com/cd/E11882_01/appdev.112/e41502/adfns_cqn.htm#ADFNS
>>> 018>?
>>> -KJ
>>>
>>> Kevin Jernigan
>>> +1 650 607.0392 (office) | +1 415 710.8828 (mobile)
>>> Senior Director, Product Management, Data Layer
>>> Compression (ACO & HCC) | Resource Management (DBRM & IORM)
>>> Database Smart Flash Cache | Temporal SQL and CQN
>>> SecureFiles and filesystems (dNFS & DBFS & CloneDB)
>>> Database Development, Oracle
>>>
>>> On 8/27/2013 12:48 PM, Jeremy Schneider wrote:
>>>> Have a question for the list. What do you think would be the best way
>>>> to
>>>> provide a set of schema-specific alerts with fairly sophisticated
>>>> trigger
>>>> and notification settings? OEM (or your general network monitoring
>>>> tool)
>>>> or some custom code closer to the application?
>>>> I have several cases where I need to provide one-off alerts for things
>>>> like
>>>> "number of rows in this table exceeds threshold" and the alerts aren't
>>>> for
>>>> me personally but for other business groups. Need specific alert
>>>> messages
>>>> for "critical" and "clear" thresholds and also the ability to
>>>> re-generate
>>>> critical alerts every X hours if the condition continues. Also want to
>>>> tweak how frequently each check is run (between every 5 to 15 minutes).
>>>>
>>>> OEM provides a great framework for handling all sorts of notification
>>>> situations. I can define the SQL as a Metric Extension. However I'm
>>>> currently using Administration Groups to automatically apply generic
>>>> monitoring templates across a broad set of databases... and it seems
>>>> against this philosophy to start having custom thresholds or
>>>> notifications
>>>> on a per-database (not to mention per-schema) level. Nonetheless it
>>>> seems
>>>> that any other approach involves a degree of re-inventing the wheel
>>>> when it
>>>> comes to the alerts, thresholds, repeated notifications, etc.
>>>>
>>>> Thoughts? What's a good architecture for this - am I missing something
>>>> obvious?
>>>>
>>>> -Jeremy
>>>>
>>>>
>>>> --
>>>> http://about.me/jeremy_schneider
>>>>
>>>>
>>>> --
>>>> http://www.freelists.org/webpage/oracle-l
>>>>
>>>>
>>>
>>>
>>> --
>>> http://www.freelists.org/webpage/oracle-l
>>>
>>>
>>>
>>>
>>

>
> --
> http://www.freelists.org/webpage/oracle-l
>
> --

http://www.freelists.org/webpage/oracle-l Received on Wed Aug 28 2013 - 17:49:59 CEST

Original text of this message