If you open a case with support and you are on 6.x version and hence their first solution always will be asking the customer to upgrade to new version. I would recommend you ask them or you find out the root cause of the failure in that particular time since it is easily reproducible every day night. So that you have better idea of the issue not by simply upgrading to new version.
Generally support doesn't point a bug and suggest you to upgrade your LDAP infrastructure just for fun.
This is a public/known defect, clearly identified since years and was due to a code change in certificate handling section that in the first 6.x releases was not that optimized; however I'm afraid that the internal code (both bugged and fixed) cannot be disclosed.
A confirmation of this potential problem could be found analysing the usage pattern of the I/O subsystem around that time; do you have any
# iostat -zxcCndT d <INTERVAL>
collected around that time? I would expect an increase of the asvc_t.
Are there any clients disconnecting/reconnecting at that time?