Re: SAP and HANA

From: Mindaugas Navickas <>
Date: Fri, 18 May 2012 06:28:45 -0700 (PDT)
Message-ID: <>

I do not have hands-on experience myself, but our client had implemented HANA recently in their SAP environment and they are pretty happy with initial results - query that was running 5 minutes 30 seconds (SAP+Oracle11.2), in SAP+HANA was completing in 12 seconds. Where all speed of HANA comes from comparing with traditional RDBMS? Different data storage concept - Column-oriented storage versus Row-oriented storage which allows fast access to data without using indexes, good compression ratio due to repeating data in columns - good academical paper on Column-oriented storage is here This is In-memory database. Please do not mix this concept with In-memory data caching - more reading on that, My understanding is that similar concepts are embeded into Microsoft PowerPivot engine ( At this point HANA is read-only database (no transactions allowed) - in next version SAP claims that it will support transactions. I am just sharing information I have - it's up to you to interpret/decide if those new technologies have any future in DB landscape.  
Mike Navickas
Oracle&DB2 DBA
From: Subodh Deshpande <> To:
Cc: oracle-l <> Sent: Friday, May 18, 2012 4:30:21 AM
Subject: Re: SAP and HANA
The way in which new technologies are emerging such kind of challenges may arrive
If a revolutionary change occurs and if a new database can replace DB2 and/or Oracle, then definitely it will be a matter of concern.

Mostly HANNA will require more technical resources to manage the architecture, when every one is cutting down the cost on their present assets.

primafacie in memory database in this new sap and oracle times ten database are not different, and now please tell me the difference between times ten and huge sga with almost all objects cached in memory ! I think such databases are suitable for dataware house where as SAP is known for process means the databases will be more General purpose or it will benefit..i think somewhere it is written new sap db is going to use rule based optimiser

if suppose tomorrow a new features emerges in a database technology and in SQL engine, like when a query is fired it will decide where its a pure select query or also has dmls, ddls in its other query portion, it will decide what should be its plan, depending upon this the query execution will change..the nature of usage of undo, buffer cache will change...means memory we just mark dirty buffers..if dirty buffers are flagged depending upon the type of query the execution will change..may there will be multiple lru lists

by flagging I mean, select and insert requires index position, where as update and delete may change the index position..the query when fired should decide what is the nature of the query and should go ahead..

these are just thoughts..may be wrong..I am just a database user working as dba...

thanks and take care..subodh

On 17 May 2012 09:07, Art Vandelay <> wrote:

> List,
> We're seeing some discussions in our organization around the SAP
> Database offering Hana.
> I'm interested to hear your thoughts on this
> Do you think this product will gain any traction?
> Do you believe Gartner when they state "A Double Wake Up Call for
> SAP's Competitors" - Should Larry be worried?
> Is anyone running Hana and can they provide some feedback?
> I appreciate your thoughts on the matter.
> Steve
> --

Love me or Hate me both are in my Favour.
Love me, I am in your Heart. Hate me, I am in your Mind.

Received on Fri May 18 2012 - 08:28:45 CDT

Original text of this message