Hi,
Le Lun 5 mai 2008 23:08, Howard Chu a écrit :
The logs show that the syncrepl consumers all timed out periodically, when trying to bind to a provider. It seems that using a 1 second timeout in the syncrepl configs is too short, or your test machine was too slow during that run.
Probably we should remove that timeout now, since the cn=config/thread pause issue has already been resolved.
Currently testing new RE24. I have now a slapd process stuck at 100% : _ /bin/sh ./scripts/test050-syncrepl-multimaster hdb _ /tmp/openldap/tests/../servers/slapd/slapd -s0 -F slapd.d -h ldap://localhost:9011/ -d 261 _ /tmp/openldap/tests/../servers/slapd/slapd -s0 -F ./slapd.d -h ldap://localhost:9012/ -d 261 _ /tmp/openldap/tests/../servers/slapd/slapd -s0 -F ./slapd.d -h ldap://localhost:9013/ -d 261 _ /bin/sh ./scripts/test050-syncrepl-multimaster hdb _ ../clients/tools/ldapsearch -P 3 -x -LLL -H ldap://localhost:9013/ -s base -b cn=Ursula Hampster,ou=Alumni Association,ou=People,dc=example,dc=com (objectClass=*) _ awk /^dn:/ {print "OK"}
Do you need some debugging info (strace, gdb, log, or other) ?
Regards, Raphael Ouazana.