Re: LDAP v. RDBMS for large datasets

[Giovanni Baruzzi <giovanni.baruzzi@allianz-leben.de>]

Hallo All,

My opinion is that LDAP may be equal on a single query but much faster on
multiple: the reason is that LDAP doesn't need the paraphernalia of a normal
SQL Server, because it doesn't need to support transactions.
In other words: a SQL server has to implement journaling in order to separate the
single transaction. Journaling can be extremely expensive.
LDAP doesn't support transaction integrity and its need for journaling are far
less (or almost inexistent). Having more resources free, LDAP can the be faster.

bye
Giovanni


Paul Makepeace wrote:

> I'm curious whether anyone has any opinion/evidence of whether a fixed-length
> query match "SELECT uid=xxx" would be quicker (generally, or in specific
> cases) versus (uid=xxx) for large ranges of xxx, in the hundreds of millions.
> No other properties of either SQL or LDAP search filters are required. Assume
> LDAP-friendly high read:write ratio, no transactions, etc.
>
> Sorry to be this vague, I'm really just after ball-park estimates, i.e one is
> drastically quicker than the other or they're about the same or perhaps even
> odd-OS/platform specific situations. Touchy-feely answers are fine :-)
>
> Thanks,
> Paul
>
> --
> Paul.Makepeace@realprogrammers.com
> Thus spake the Master Programmer:
>   "Let the programmers be many and the managers few --
>     then all will
>       be productive." (http://misspiggy.gsfc.nasa.gov/tao.html)

begin:vcard 
n:Dr. Baruzzi;Giovanni
tel;work:+49-711-663-1421
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:giovanni.baruzzi@allianz-leben.de
fn:Giovanni Dr. Baruzzi
end:vcard