Hi Mladen,
First of all thanks for the response.
The primary reason they are considering this migration is to attain a
better performing database considering the scaling capabilities of Aurora
would solve all the issues which makes a technological decision not a
business decision.
My disagreement is about the idea of scaling big would fix all the
problems.
This is what I think about our scenario:
"If Camaro is not fast enough for you, definitely moving to a 18 Wheeler
will not help". I might be completely wrong here.
Thanks again for the detailed comparison. :)
On Thu, Apr 5, 2018 at 4:46 PM, Mladen Gogala <gogala.mladen@xxxxxxxxx>
wrote:
Hi Ravi,
First, the move from Oracle to MySQL (or Aurora) is a business decision,
not a technological decision. Such move is usually being caused by the
cost, not by technological merits of one database over the other database.
There are several things about Oracle pricing that you must have in mind:
- Oracle is the only RDBMS vendor which charges for the right to
create another database in the instance.
- Oracle is the only RDBMS vendor which charges for the right to
create a partitioned table.
- Oracle is the only RDBMS vendor which charges for the right to
maintain columnar and row storage simultaneously. MSSQL and DB2 both have
this capability, but is included into enterprise edition license. MariaDB
offers the same capability for free.
Also, if your company has undergone an Oracle auditing, then the decision
to abandon Oracle makes even more sense. Oracle auditing event can be
compared with the "Is it safe?" scene from the movie "Marathon Man" or much
more famous "squealing" scene from the movie "Deliverance".
As for the benchmark part of your question, you are trying to compare
Chevy Camaro with an 18-wheeler. Is Camaro faster? It certainly is. Can
you transport 40 tons of stuff in a Camaro? Probably not. It all depends on
your business needs. How are you using your database? I would probably have
a problem with any version of MySQL being used as a mixed mode
OLTP/reporting database which needs to support 5000 simultaneous online
users, with 99.99% availability requirement. Those requirements are fairly
modest and correspond to medium size insurance company or an HMO. Major
banks, major retailers like Walmart, Target, CostCo or Amazon,
communication companies like AT&T, Time-Warner, Charter and Verizon or
credit card companies like American Express or Visa Inc. all have much more
than 5000 simultaneous users. All of them have databases which need to be
up 99.99% of the time, 365 days per year, 7 days a week, 24 hours a day.
Some of their databases are gigantic, 100's of TB. We are probably on the
verge of the 1st PB database at one of those companies. Only Oracle and DB2
can do that.
On the other hand, if your company is relatively small, with the database
being used to power a website, visited by around 50 simultaneous users at
any given moment, you may get by with some variant of MySQL. Long story
short, it all depends on your business requirements. Benchmarks are
irrelevant, TPC went down the drain when Oracle withdrew from the
organization. There is no relevant benchmark for databases. It seems that
in your case, the business decision has already been made. If you disagree
with it, look for another job. It's the only honest thing to do. The fact
that you are against that decision is completely irrelevant. You have to
think about your career: do you want to become an Aurora DBA? You will
probably have an opportunity not only to learn Aurora, but to become AWS
certified person, which can look pretty good on your resume. Do you want to
remain Oracle DBA? That is a legacy job which has always paid well,
although it's getting harder by the day to find a good Oracle-only job.
It's about what you want. Your company has already made a business
decision. Now, it's your turn.
On 04/05/2018 04:09 PM, Ravi Teja Bellamkonda wrote:
Hi List,
My organization is planning on moving from Oracle to Aurora MySQL
(capability of having Read Replicas) for the sake of performance and
scalability which I am completely against as I believe that having the
capability of scaling will not solve all the problems.
I cannot find any benchmarks comparing Oracle with MySQL. Can someone
please provide insights on whether this is a even a good move as I am new
to MySQL and not sure how good it is in terms of performance when compared
with Oracle. As far as I know Oracle is better performant.
Your time is appreciated.
--
Thanks & Regards,
Ravi Teja
--
Mladen Gogala
Database Consultant
Tel: (347) 321-1217
