[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: Another try (scalability, usage)



========> On 2000-07-26, Adam Haberlach <Adam> wrote:

    Adam> o. Is anyone using OpenLDAP for large (10,000 or so)
    Adam>    directories?  Does it work well at this level?  How
    Adam>    heirarchial is your schema?  We're looking at a situation
    Adam>    in which most users will be in one organizational unit
    Adam>    (mainly for authentication).

At the moment we have approx. 80,000 entries in our OpenLDAP
directory. It works very well and I can't see any performance
problems, at least for searches over indexed attributes. Insertion and
updates are not performance critical for our application.

The directory tree consists of six levels and the majority of our
entries is sitting at the bottom level.

    Adam> o. Is anyone storing arbitrary data?  We are also looking at
    Adam>    storing alternate versions of filesystems, for revision
    Adam>    tracking.

The above directory holds bibliographic references -- you can call
this arbitrary data in comparison with users and account information.

-- 
Jens Klöcker <kloecker@springer.de> | There is one art, no more, no less:
Electronic Technologies             | to do all things with art-lessness
Springer-Verlag                     | -- Piet Hein