I have been investigating high cpu utilization on our SLES10SP3/OES2SP2 server that is the user source for our ZCM 10SP3 system.

Using ndstrace on the ldap filter, I am noticing a continuous string of errors like the following:

1322916160 LDAP: Illegal ndsname "Jenna_Opperman" in ldap2uNDSDN, err = 34 (0x22)
1322916160 LDAP: ldap2uNDSDN ldapDN = "Jenna_Opperman" - error 34 (0x22)
1322916160 LDAP: Failed to convert LDAP DN "Jenna_Opperman" in nds_back_bind, err = 34 (0x22)
1225247040 LDAP: Connection 0x6d569c0 read failure, setting err = -5874

The errors we receive are all from different users and we receive them continuously. The cpu of hthe server spikes up to 60+% regularly while other replica masters rarely hit 10%. There is no difference between the ldap group and server objects for the high and normal cpu util servers.

Obviously "Jenna_Opperman" is not a valid ldap DN, but why is the system receiving this form of the name? The odd thing is that we're not receiving reports of obvious issues from these users. We can browse the user source fine and zcm services work normally for the most part. However, I'm concerned about the high utilization and the potential for the issue to scale into a serious issue.

I've seen other posts that referred to this message after a disaster recovery situation, but we have not done anything like that. I've also seen speculation that CASA may be involved in the passing of this bad DN format.

Any suggestions on how to further troubleshoot this issue?