Re: youtube video on pgsql integrity

From: Zdeněk Bělehrádek <zdenek(dot)belehradek(at)superhosting(dot)cz>
To: pgsql-general(at)postgresql(dot)org
Subject: Re: youtube video on pgsql integrity
Date: 2012-11-29 16:32:13
Message-ID: op.wojmvzcpgvy402@localhost.localdomain
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Dne Thu, 29 Nov 2012 16:00:55 +0100 Ray Stell <stellr(at)vt(dot)edu> napsal(a):

>
> On Nov 29, 2012, at 9:27 AM, Kevin Grittner wrote:
>> is everything shown there really
>> the behavior of the MySQL database itself?
>
> Good question. I intend to install mysql one day to explore, but just
> can't find the time. The particular engine is not disclosed and I've
> read some are better than others....
>

It is behaviour of SQL parser, independent of storage engine. I tried it
with MySQL in default settings, results are at http://pastebin.com/mbavfdj9
As you can see, your data will be changed on both mainstream engines.

There is some "funny" effects created by this architecture (SQL parser and
multiple storage engines), my favourite are the ones when InnoDB refuses
to start (e. g. when you change innodb_log_file_size and forget to erase
old innodb logs). Rest of MySQL starts happily, and even lets you drop
tables belonging to unknown engine (InnoDB). When you realize your mistake
and restart MySQL with functional InnoDB, you won't see the table (because
SQL metadata are gone), but you can't create it either (because table
metadata are still stored in InnoDB tablespace).

-- Zdeněk Bělehrádek

In response to

Browse pgsql-general by date

  From Date Subject
Next Message icholy 2012-11-29 16:40:06 libpq - PQsendQuery wait for complete result
Previous Message Ryan Kelly 2012-11-29 16:15:42 Re: When does CLUSTER occur?