atxl wrote: > On Tuesday, January 21, 2003, at 05:57 PM, Nico Erfurth wrote:
>
>> atxl wrote:
>>
>>> Okay, after further discussion and testing, it turns out that the LDAP
>>> lookup is _not_ the culprit. Any suggestions as to where to look to
>>> find something which could cause these symptoms?:
>>>
>>> Multiple recipients result in the first listed recipient being properly
>>> delivered, the others go into permanent loop with 'local_domains lookup
>>> deferred' messages, even when there is no lookup for local_domains (I
>>> replaced the lookup with a colon seperated list of the local domains
>>> for the box.)
>>
>>
>> Maybe a DNS-issue?
>>
>> Run exim with exim -d9 and show us the relevant parts of the logfile,
>> this helps much more.
>>
>
> Ok, spent a few hours' quality time with the voluminous (and extremely
> informative) d9 logs, found the following at the end of the output for
> a particular message:
>
> ---Trimmed at beginning of seemingly salient part ---
>
> Re-using cached connection to LDAP server 216.174.10.25:0
> Start search
> search ended by ldap_result yielding -1
> ldap_result failed
> exim: /mnt/rhome/torsten/packages/openldap2/openldap2-2.0.27/libraries/
> libldap/error.c:221: ldap_parse_result: Assertion `r != ((void *)0)'
> failed.
>
> --- ---
>
> This occurs when exim is attempting to do local delivery for the second
> recipient. Processing stops at this crash. The first recipient has
> already been completely processed and successfully delivered. This
> recipient has been processed most of the way thru, and this ldap search
> is to determine the user to deliver as (to determine the uid for
> delivery.)
>
> Anyone seen this before? Suggestions?
This looks like an openldap bug, did you compile openldap yourself?
Please try to compile exim against THIS version.
