Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: David Wilson <david(dot)t(dot)wilson(at)gmail(dot)com>
Cc: Marko Kreen <markokr(at)gmail(dot)com>, Alvaro Herrera <alvherre(at)commandprompt(dot)com>, toruvinn <toruvinn(at)lain(dot)pl>, pgsql-bugs(at)postgresql(dot)org, pgsql-performance(at)postgresql(dot)org
Subject: Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance
Date: 2009-07-15 15:30:26
Message-ID: 23164.1247671826@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs pgsql-performance

David Wilson <david(dot)t(dot)wilson(at)gmail(dot)com> writes:
> On Wed, Jul 15, 2009 at 11:10 AM, Marko Kreen<markokr(at)gmail(dot)com> wrote:
>> From security standpoint, wasting more cycles on bad passwords is good,
>> as it decreases the rate bruteforce password scanning can happen.
>>
>> And I cannot imagine a scenario where performance on invalid logins
>> can be relevant..

> DoS attacks. The longer it takes to reject an invalid login, the fewer
> invalid login attempts it takes to DoS the server.

Yeah, but even with the current setup, an attacker who can fire
connection request packets at your postmaster port is not going to have
any trouble DoS'ing the service. We expend quite a lot of cycles before
getting to the password challenge already.

regards, tom lane

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Marko Kreen 2009-07-15 15:31:25 Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance
Previous Message Alan Pinstein 2009-07-15 15:27:58 Re: BUG #4921: ltree @> ltree[] operator shouldn't fail if ltree[] is empty

Browse pgsql-performance by date

  From Date Subject
Next Message Marko Kreen 2009-07-15 15:31:25 Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance
Previous Message David Wilson 2009-07-15 15:16:23 Re: [PERFORM] BUG #4919: CREATE USER command slows down system performance