Re: MYSQL Error 2013 load infile 15mln rec 6gb CSV

From: Mladen Gogala <gogala.mladen_at_gmail.com>
Date: Mon, 20 Oct 2014 16:19:22 +0000 (UTC)
Message-ID: <pan.2014.10.20.16.19.22_at_gmail.com>


On Sun, 19 Oct 2014 20:57:40 +1100, Noons wrote:

> Ah well, I'm quite sure that loads of austerity, lots of "cloud spray"
> and all sorts of other inane campaigns will address all that!
> Real soon now.

Oracle's good fortune is that none of the Oracle's competitors is in good shape right now. One competitor is desperately trying to establish its presence in the tablet and smart phone market while neglecting everything else , and throwing more than $1B annually on the still failing Surface. Another competitor is losing money for 9 quarters in a row and selling off its business units to Lenovo. They too believe in the cloud pixie dust.
Open source world is very far from offering functionality that could make it viable. There are two principal open source databases and one is fragmented into Oracle owned piece, MariaDB and Drizzle and the other one is managed by idiots who don't see the usefulness of hints. Hints are just a symptom of what is wrong with Postgres. Here is a sample of the mentality among the Postgres designers:

http://it.toolbox.com/blogs/database-soup/why-postgresql-doesnt-have-query-hints-44121?rss=1

So, the state of Oracle competition is really desperate. As for the Exadata failing, it's because that is one area where viable competitors do exist. I'm talking about EMC Greenplum, of course. Greenplum is faster and cheaper than Exadata and it's easier to use. Exadata is a half baked "smart storage" which essentially speeds up full table scans. Greenplum is a DW database which uses columnar store and is a full fledged DW database, a turnkey solution which doesn't really need additional EE and RAC licenses to work. That may have something to do with a former Exadata performance architect who betrayed Oracle Corp. and deserted to EMC? He was immediately stripped off his Oracle Ace title. I'm sure he's crying inconsolably about that. That is the reason why I never bothered to learn Exadata. Long story short, Oracle still has a chance, because of the lackluster competition.

-- 
Mladen Gogala
The Oracle Whisperer
http://mgogala.byethost5.com
Received on Mon Oct 20 2014 - 18:19:22 CEST

Original text of this message