On Wed, Aug 07, 2002 at 09:31:50AM -0400, CHENNES@Nationwide.com wrote:
I am having difficulty keeping 1.5.19 running when receiving syslog messages via tcp. I have a central logging host that receives messages via tcp from collection hosts. The collection hosts are receiving messages from various hosts via udp and tcp. Essentially, the ng process immediately dies on the central logging host as soon as I start it. However, if I alter the configurations to use only udp, I have no problems. The central logging host and the collection hosts are all running Solaris 8 with 1.5.19. I have tested the same exact scenario using a previous version of ng (1.4.14) on the central logging host and it works just fine. I have included some configuration information below.
can you provide a gdb backtrace based on the core file? gdb syslog-ng -c core gdb> bt and the results. -- Bazsi PGP info: KeyID 9AF8D0A9 Fingerprint CD27 CFB0 802C 0944 9CFD 804E C82C 8EB1