OK, I got back to the rig and had a play with it again (I can
reproduce this error on another installation too - seems only
to affect large directory fetches and sometimes can be avoided
by using a small 'page' size ( say 5 entries ).
All libraries are up to date.
The SSLeay trace level was set to 4 (dump data). Unfortunately the
dump is in binary not text and it is difficult to correlate this
with the LDAP debug output. I will have a play and add some more
diagnostics to the code to see if I can get to the bottom of this
( 300Mb log files on a VM do not make for lightening fast diagnostics :-) ).
Si
Chris Ridd wrote:
On 19/5/06 1:44, Simon P Smith <[EMAIL PROTECTED]> wrote:
Chris Ridd wrote:
On 15/5/06 3:26, Si Smith <[EMAIL PROTECTED]> wrote:
So instead of starting "30 84 00 00" the bad record
starts "00 84 00 00".
Does this make any sense?
It does, and it explains the LDAP_DECODING_ERROR. Now, why is it
happening...
What version of Convert::ASN1 are you using?
Perl 5.8.6 ( ActivState Win32 811 on one machine 816 on another )
Convert::ASN1 Version 0.18 ( CVS 1.29 )
(ahh, I see that there is a ppd for 0.20 I guess this is
a recommended action before carrying on?)
Yeah, there were some bugs fixed there relating to UTF-8 and 'bytes', which
*could* be what you're seeing. Graham's suggestion of getting Net::SSLeay's
idea of the bytes is an excellent one too.
Cheers,
Chris
