What does etime for BIND operation really indicate?
807573Feb 17 2009 — edited Mar 3 2009Hello all -
Running 5.2 P4 on Solaris 10. We have a client application which will continuously open up a new LDAPS connection, BIND with a service account, search, then disconnect. Yes - very inefficient. We have noticed "high" etimes on the BIND operations but it seems the etime actually starts from somewhere within the LDAPS handshake - not the actual BIND operation itself. Please see below:
[17/Feb/2009:14:28:38 -0500] conn=83754 op=-1 msgId=-1 - fd=21 slot=21 LDAPS connection from xx.xxx.xx.xxx to yy.yyy.yy.yy
[17/Feb/2009:14:28:41 -0500] conn=83754 op=-1 msgId=-1 - SSL 128-bit RC4
[17/Feb/2009:14:28:41 -0500] conn=83754 op=0 msgId=1 - BIND dn="binddn" method=128 version=2
[17/Feb/2009:14:28:41 -0500] conn=83754 op=0 msgId=1 - RESULT err=0 tag=97 nentries=0 etime=3 dn="binddn"
Notice the etime=3 on the result of the bind operation references with the LDAPS connection not the the "SSL 128-bit" line or the BIND. Is this expected behavior? Any information on this process would be helpful
Thanks
Randy
Edited by: RandyWatson on Feb 17, 2009 5:24 PM