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

Re: OpenLdap's Scalability

Hello all

I have no answer for the moment, but as I am testing the scalability of
OpenLDAP too, I encountered a problem :
I generated a LDIF file with more than one million entries (500 users,
2500 groups of  users, 1 250 000 groups of groups).
As my OpenLDAP can add about ten entries per second with ldapadd, I
thought it would be a good idea not to spend 125 000 seconds = 35 hours
adding my data just for a test, and used slapadd.
so I stopped the server, removed all the /var/lib/ldap/* files, and ran
slapadd with -b "dc=my,dc=suffix" -l myfile.ldif  .
The day after, I started slapd, and did a search on the suffix, which
returned "no such object".

slapcat do tell me the suffix exists in the database
I use the ldbm backend as it is the default one in Red Hat Entreprise
Server 3.0

Did I make a mistake, or did I find an inconsistency of slapd/slapadd/ldbm ?


Kurt Palmer a écrit :

Hello All,
I am trying to assess the scalability of OpenLdap for a project i am working on. I am aware that Sun's Iplanet is capable of scaling to millions of users and would like to know [especially to prove to upper management] if OpenLdap is being "used" in such a large scale deployment.
Thanks in Advance -Kurt

Kurt Palmer
Systems Engineer
Vonage Holdings
2147 RT 27
Edison NJ,08817