Skip site navigation (1) Skip section navigation (2)

Re: BLOBs

From: "Chad Thompson" <chad(at)weblinkservices(dot)com>
To: <aagha(at)bigfoot(dot)com>, "Warwick Hunter" <whunter(at)agile(dot)tv>
Cc: "pgsql-novice" <pgsql-novice(at)postgresql(dot)org>
Subject: Re: BLOBs
Date: 2002-08-16 16:52:53
Message-ID: 030601c24545$722989a0$32021aac@chad (view raw or flat)
Thread:
Lists: pgsql-novice
That is exactly what i have been doing to date.  I have always thought that
it would be cool to manage these files (.wav in my case) in a database, w/o
the problem of some user mucking with a directory and screwing up the files.

I believe that i'll give it try.

Thanks for all your responses
Chad
----- Original Message -----
From: "Aurangzeb M. Agha" <aagha(at)bigfoot(dot)com>
To: "Warwick Hunter" <whunter(at)agile(dot)tv>
Cc: "Chad Thompson" <chad(at)weblinkservices(dot)com>; "pgsql-novice"
<pgsql-novice(at)postgresql(dot)org>
Sent: Thursday, August 15, 2002 10:26 PM
Subject: Re: [NOVICE] BLOBs


> I hope this isn't getting too academic, but why not store the path to
> blobs in the DB and access them in some other fashion?
>
> Rgs,
> Aurangzeb
>
> On Fri, 16 Aug 2002, Warwick Hunter wrote:
>
> :Chad
> :
> :I am using PostgreSQL to store BLOBS. I can't say that the
> :performance I am getting is wonderful, but it is adequate
> :for my purposes for now. Few people are every satisfied
> :with the performance of their database anyway :-)
> :
> :I have done some performance testing on PostgreSQL and have
> :discovered that on an Linux Intel machine with reasonably fast
> :SCSI discs I can get average performance numbers like these:
> :
> :Binary BLOB Size
> :300 KB read=24  ms write=185 ms del=105 ms
> :150 KB read=11  ms write=90  ms del=51  ms
> :   5 KB read=2.5 ms write=14  ms del=11  ms
> :
> :I have done a small amount of tuning of the PostgreSQL
> :configuration. I plan to do some more.
> :
> :> Or is storing BLOBs in a DB a good idea?
> :
> :I have found PostgreSQL to be quite robust. I have a
> :test that deliberately kills the postmaster and postgres
> :subprocesses while doing write operations to see if I
> :can cause the database to be corrupted. This test
> :rarely fails.
> :
> :This test fails more regularly if you rewrite the
> :data of an existing BLOB rather than create a new
> :one and delete the old one. So now I only ever create
> :a new BLOB.
> :
> :I have seen one instance when the power on a machine
> :failed while writing a BLOB caused one BLOB to not
> :be restored correctly when the database came up.
> :I plan to investigate this and report a bug if
> :appropriate.
> :
> :I looked at the bytea data type for storing my objects
> :as regular columns. However I was unable to make it work
> :with the normal SQL statements, the binary data always
> :upset the SQL command and the escaping routines didn't
> :seem to help.
> :
> :Warwick
> :
>
> --
> Aurangzeb M. Agha    | Email : aagha(at)bigfoot(dot)com
>                      | Home  : +3 010 8959.558
> 34 Nafsika St.       | Direct: +3 010 8995.875
> Voula 16673          | Mobile: TBD
> Greece               | Voice : 415 412 4234 (US Only)
>                      | e-Fax : 978 246.0770
> PGP ID: 0x68B3A763   |
>
> "Those who would give up essential liberty to purchase a little
>  temporary safety deserve neither liberty nor safety."
>
> - Benjamin Franklin
>
>


In response to

  • Re: BLOBs at 2002-08-16 04:26:27 from Aurangzeb M. Agha

Responses

  • Re: BLOBs at 2002-08-16 18:53:37 from Aarni Ruuhimäki / Megative Tmi / KYMI.com

pgsql-novice by date

Next:From: Chad ThompsonDate: 2002-08-16 17:47:06
Subject: No space left on device.
Previous:From: John TaylorDate: 2002-08-16 16:27:07
Subject: Re: COPY FROM with a serial not null column

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group