[Pkg-openldap-devel] Bug#798964: Bug#798964: slapd: since security upgrade writing ber in accesslog is broken

[Ryan Tandy]

Control: tag -1 moreinfo

Hi Benoit, thanks for reporting this.

On Mon, Sep 14, 2015 at 04:37:11PM +0200, Benoit Mortier wrote:
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 fd=25 ACCEPT from
>IP=172.20.51.10:51214 (IP=172.20.50.25:389)
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 op=0 BIND dn="cn=o3-
>synchro,ou=dsa,dc=master,dc=fdi" method=128
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 op=0 BIND dn="cn=o3-
>synchro,ou=dsa,dc=master,dc=fdi" mech=SIMPLE ssf=0
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 op=0 RESULT tag=97 err=0
>text=
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 op=1 SRCH
>base="cn=accesslog" scope=2 deref=0
>filter="(&(objectClass=auditWriteObject)(reqResult=0))"
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 op=1 SRCH attr=reqDN
>reqType reqMod reqNewRDN reqDeleteOldRDN reqNewSuperior entryCSN
>Sep 14 16:28:25 agenor slapd[11359]: send_search_entry: conn 1000  ber
>write failed.
>Sep 14 16:28:25 agenor slapd[11359]: conn=1000 fd=25 closed (connection
>lost on write)

I'm not seeing any such errors in a delta-syncrepl setup. Can you share 
the relevant configs, and possibly the corresponding logs from the 
consumer? It might be necessary to capture some logs at a higher debug 
level (for example "stats sync trace") to figure out what's going on.

If you downgrade to the previous version of slapd, does the problem go 
away?

thanks,
Ryan