Date: Fri, 15 Dec 2000 11:35:14 +0000
From: Mark Robinson
Message-ID: <5392612389.20001215113514@reason-technology.com>
Subject: DNS/Firewall/oops problem
Hi all,
I'm having trouble with named/firewalling/ISDN. Basically, what
happens is that it dials up, something seems to crash, and the
firewall won't recognise the DNS server. firewals then blocks port
53, and high dns ports, but the ISP's nameservers keep trying to
reply, and keep the isdn line up forever!
I don't know anything about the Oops or stack trace stuff. Is it
possible to tell what happened?
Thanks in advance
Mark
----------------------------------------------------------------------------------
DNS request for something...
Dec 12 23:57:48 rmdserve kernel: OPEN: y.y.y.y -> x.x.x.x UDP, port: 3532 -> 53
Dec 12 23:57:48 rmdserve kernel: ippp0: dialing 1 <ISPTEL>...
[snip]
Dec 12 23:57:49 rmdserve ipppd[207]: Local number: 0<TEL>1, Remote number: <ISPTEL>, Type: outgoing
Dec 12 23:57:49 rmdserve ipppd[207]: PHASE_WAIT -> PHASE_ESTABLISHED, ifunit: 0, linkunit: 0, fd: 7
Dec 12 23:57:49 rmdserve kernel: isdn_net: ippp0 connected
Dec 12 23:57:49 rmdserve ipppd[207]: Remote message: <HOST>: IP Address: y.y.y.y Running PPP on 2264
Dec 12 23:57:49 rmdserve ipppd[207]: MPPP negotiation, He: No We: No
Connected ok, but...
Dec 12 23:57:50 rmdserve kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000008
Dec 12 23:57:50 rmdserve kernel: current->tss.cr3 = 02d6e000, %cr3 = 02d6e000
Dec 12 23:57:50 rmdserve kernel: *pde = 00000000
Dec 12 23:57:50 rmdserve kernel: Oops: 0000
Dec 12 23:57:50 rmdserve kernel: CPU: 0
Dec 12 23:57:50 rmdserve kernel: EIP: 0010:[lock_get_status+29/280]
Dec 12 23:57:50 rmdserve kernel: EFLAGS: 00010286
Dec 12 23:57:50 rmdserve kernel: eax: 00000000 ebx: c02a8de0 ecx: 00000002 edx: c023007f
Dec 12 23:57:50 rmdserve kernel: esi: c05ea660 edi: c2aad05d ebp: c3b2df00 esp: c3b2debc
Dec 12 23:57:50 rmdserve kernel: ds: 0018 es: 0018 ss: 0018
Dec 12 23:57:50 rmdserve kernel: Process lsof (pid: 10328, process nr: 68, stackpage=c3b2d000)
Dec 12 23:57:50 rmdserve kernel: Stack: c2aad05d c011cc86 c1d3df40 40014000 c3b2def0 c0132e2f c05ea660 00000002
Dec 12 23:57:50 rmdserve kernel: c023007f 00001000 00000000 c3b2df80 c0279920 00000400 0000005d 00000002
Dec 12 23:57:50 rmdserve kernel: c05ea660 c2aad05d c1d3df40 40014000 40015000 c0157bf9 c2aad000 c3b2df80
Dec 12 23:57:50 rmdserve kernel: Call Trace: [merge_segments+30/404] [get_locks_status+79/400] [tvecs+14367/14432] [get_root_array+449/484] [array_read+225/456] [array_read+46/456] [sys_read+178/208]
Dec 12 23:57:51 rmdserve kernel: [error_code+45/64] [system_call+52/64]
Dec 12 23:57:51 rmdserve kernel: Code: 8b 78 08 52 51 68 ed ff 22 c0 53 e8 57 32 0f 00 01 c3 83 c4
Dec 12 23:57:52 rmdserve ip-up: Warning: FW_SERVICE_DNS defined, but no DNS server found running!
Dec 12 23:57:52 rmdserve SuSEfirewall: Firewall rules successfully set.
ps -e shows named running, but it doesn't respond to nslookup etc...
Dec 12 23:57:53 rmdserve kernel: Packet log: input DENY ippp0 PROTO=17 x.x.x.x:53 y.y.y.y:3532 L=352 S=0x00 I=63711 F=0x0000 T=61 (#139)
Repeat for 9 hours!
This also appears in /var/log/messages. Is this isdn related???
Dec 13 00:00:15 rmdserve kernel: iput: inode 03:08/518337 still has locks!
Dec 13 00:00:23 rmdserve kernel: iput: inode 03:06/150 still has locks!
Dec 13 00:00:23 rmdserve kernel: iput: inode 03:06/13635 still has locks!
Dec 13 00:00:30 rmdserve kernel: iput: inode 03:08/81040 still has locks!
But BIND seems to be running!
Dec 13 00:33:12 rmdserve named[557]: Cleaned cache of 1 RRset
Dec 13 00:33:12 rmdserve named[557]: USAGE 976667592 975524994 CPU=7.3u/4.27s CHILDCPU=0u/0s
Dec 13 00:33:12 rmdserve named[557]: NSTATS 976667592 975524994 A=963 PTR=30530 MX=416 AAAA=469 ANY=1210
Dec 13 00:33:12 rmdserve named[557]: XSTATS 976667592 975524994 RR=1932 RNXD=840 RFwdR=1182 RDupR=214 RFail=45 RFErr=0 RErr=0 RAXFR=0 RLame=0 ROpts=0 SSysQ=487 SAns=25408 SFwdQ=4348 SDupQ=8051 SErr=113 RQ=3358
