Do you have any logs of other segmentation faults/ecc or edac errors in /var/log/messages? This could be a bad RAM issue. - Dustin -----Original Message----- From: redhat-list-bounces@xxxxxxxxxx [mailto:redhat-list-bounces@xxxxxxxxxx] On Behalf Of Peter Shulkin Sent: Wednesday, March 03, 2010 3:33 PM To: redhat-list@xxxxxxxxxx Subject: racoon segmentation fault I finally got IPSec working between one linux server and 2 windows servers, but about every 8 hours, I get a segmentation fault and racoon goes away. The error is always: Mar 3 14:20:31 STORE191 racoon: NOTIFY: the packet is retransmitted by 128.xx.xx.xxx[500]. Mar 3 14:20:31 STORE191 racoon: WARNING: the packet retransmitted in a short time from 128.xx.xx.xx[500] Mar 3 14:20:31 STORE191 racoon: NOTIFY: the packet is retransmitted by 128.xx.xx.xx[500]. Mar 3 14:20:31 STORE191 racoon: ERROR: wrong state 8. Mar 3 14:20:31 STORE191 racoon: ERROR: failed to pre-process packet. Mar 3 14:20:31 STORE191 racoon: INFO: IPsec-SA established: ESP/Transport 128.xx.xx.xx[0]->128.yy.yy.yy[0] spi=58532716(0x37d236c) Mar 3 14:20:31 STORE191 kernel: racoon[18824]: segfault at 0000000000000018 rip 00000000004068a7 rsp 00007fff8fb715f0 error 4 where xx is one of the 2 windows servers, and yy is the linux server. This is a test setup, so not very heavy traffic right now, but there is one NFS mounted drive from the linux server to both windows servers. Racoon restarts easily, and I just have to log onto each windows server when I restart, and ping the linux, to re-establish connection. I have a ping from the linux server to each of the windows server every 5-10 minutes to keep the tunnel alive, otherwise it drops the connection. Any ideas would be appreciated. Thanks, Peter Shulkin -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list -- redhat-list mailing list unsubscribe mailto:redhat-list-request@xxxxxxxxxx?subject=unsubscribe https://www.redhat.com/mailman/listinfo/redhat-list
