[j-nsp] IPv6 causes mgd crash

Eric Van Tol eric at atlantech.net
Thu Jul 3 08:16:43 EDT 2008


Hi all,
I'm having a problem with one of our M10i's running 8.4.R1.13.  Whenever I attempt to configure certain IPv6-related statements, whether they be addresses on interfaces or certain policy-options, I immediately get kicked out of the router when trying to commit.  The following logs are in the router afterwards:

Jul  1 06:40:22.104  r1 /kernel: BAD_PAGE_FAULT: pid 79975 (mgd), uid 0: pc 0x8142464 got a write fault at 0x91205000, x86 fault flags = 0x6
Jul  1 06:40:22.104  r1 /kernel: Trapframe Register Dump:
Jul  1 06:40:22.104  r1 /kernel:  eax: 91204000   ecx: 91000000   edx: 91000000   ebx: 00001000
Jul  1 06:40:22.104  r1 /kernel:  esp: bfbda100   ebp: bfbda118   esi: 91205000   edi: 91000000
Jul  1 06:40:22.104  r1 /kernel:  eip: 08142464   eflags: 00010287
Jul  1 06:40:22.104  r1 /kernel:  cs: 001f        ss: 002f        ds: bfbd002f    es: 9119002f
Jul  1 06:40:22.104  r1 /kernel:  fs: 9100002f    trapno: 0000000c        err: 00000006
Jul  1 06:40:22.104  r1 /kernel: Page table info for PC address 0x8142464: PDE = 0x423a4067, PTE = 45702425
Jul  1 06:40:22.104  r1 /kernel: Dumping 16 bytes starting at PC address 0x8142464:
Jul  1 06:40:22.104  r1 /kernel:     89 06 83 c6 04 ff 4d e8 89 b1 84 02 00 00 83 b9
Jul  1 06:40:22.304  r1 /kernel: BAD_PAGE_FAULT: pid 79975 (mgd), uid 0: pc 0x8831dd02 got a read fault at 0x91200000, x86 fault flags = 0x4
Jul  1 06:40:22.304  r1 /kernel: Trapframe Register Dump:
Jul  1 06:40:22.304  r1 /kernel:  eax: 08eb6000   ecx: 00020000   edx: 00280000   ebx: 08eb6000
Jul  1 06:40:22.304  r1 /kernel:  esp: bfbd9f1c   ebp: bfbd9f50   esi: 91200000   edi: 090b6000
Jul  1 06:40:22.304  r1 /kernel:  eip: 8831dd02   eflags: 00010206
Jul  1 06:40:22.304  r1 /kernel:  cs: 001f        ss: 002f        ds: 002f        es: 002f
Jul  1 06:40:22.304  r1 /kernel:  fs: 9100002f    trapno: 0000000c        err: 00000004
Jul  1 06:40:22.304  r1 /kernel: Page table info for PC address 0x8831dd02: PDE = 0x39224067, PTE = 3f8ae425
Jul  1 06:40:22.304  r1 /kernel: Dumping 16 bytes starting at PC address 0x8831dd02:
Jul  1 06:40:22.304  r1 /kernel:     f3 a5 89 d1 83 e1 03 f3 a4 5f 5e c3 01 cf 01 ce
Jul  1 06:40:23.904  r1 /kernel: Failed to write core file for process mgd (error 14)

Anyone seen this before?  I opened a JTAC ticket, but it's not moving as fast as I would hope and I have to assist a government client in getting their IPv6 connectivity operational.  I imagine it's a bug with this version of JUNOS, but in the event it's not, I'd hate to upgrade for no reason.

Thanks,
evt


More information about the juniper-nsp mailing list