Hi Jon, Burgess, Jon napsal(a): > I setup a pair of Fedora 17 VMs as per the "cluster from scratch" document > and initially everything was working OK. Later I was trying to set the node > IDs manually and accidently set two nodes to use the same ID. At this point > the Corosync processes on both machines segfaulted. I understand that this > is an invalid configuration but it seems dangerous that an existing cluster > of devices can be interrupted by a new device with a duplicate ID. Would > there be any interest in me raising a bugzilla entry with the backtraces > etc? > Actually, bug is very nicely reproducible so really no need to fill BZ. I will try to fix this ASAP. > On both devices the crash occurs when the code hits a NULL pointer. First > device: > > Program received signal SIGSEGV, Segmentation fault. > 0x0000000000000000 in ?? () > (gdb) bt > #0 0x0000000000000000 in ?? () > #1 0x000000000040cd7b in sync_barrier_handler (msg=0x7ffff5f53641, > nodeid=1) at sync.c:236 > > > Second device: > > Program received signal SIGSEGV, Segmentation fault. > 0x0000000000000000 in ?? () > (gdb) bt > #0 0x0000000000000000 in ?? () > #1 0x000000000040c5a7 in schedwrk_processor (context=<optimized out>) at > sync.c:551 > > > Jon > > > Regards, Honza _______________________________________________ discuss mailing list discuss@xxxxxxxxxxxx http://lists.corosync.org/mailman/listinfo/discuss
