Re: WCCP2 compile problem

From: Reuben Farrelly <reuben-squid-dev@dont-contact.us>
Date: Wed, 17 May 2006 17:04:28 +1200

On 17/05/2006 7:38 a.m., Henrik Nordstrom wrote:
> tis 2006-05-16 klockan 12:46 -0400 skrev Jeremy Hall:
>
>> I noticed that those zeroized lists would sometimes have a router id of
>> 0.0.0.0 and then either wccp would try to send to 0.0.0.0:2048, it would
>> abort wccp2 processing when it encountered it, or it would throw an
>> error. Did you try it after zeroizing the list?
>
> The 0.0.0.0 entry was always there as the first entry in the list.. now
> it's not there anymore.
>
> I don't have any WCCPv2 equipment to run a full test with, but the code
> is quite simple..
>
> Regards
> Henrik

Speaking of WCCPv2, I'm seeing this logged when I do a squid -k reconfigure:

2006/05/17 16:48:00| Beginning Validation Procedure
2006/05/17 16:48:00| Completed Validation Procedure
2006/05/17 16:48:00| Validated 35418 Entries
2006/05/17 16:48:00| store_swap_size = 1555560k
2006/05/17 16:48:01| storeLateRelease: released 0 objects
2006/05/17 16:55:30| Reconfiguring Squid Cache (version 2.6.DEVEL-CVS)...
2006/05/17 16:55:30| FD 23 Closing HTTP connection
2006/05/17 16:55:30| FD 24 Closing WCCP socket
2006/05/17 16:55:30| FD 24 Closing WCCP socket
2006/05/17 16:55:30| assertion failed: comm.c:647: "F->flags.open"
2006/05/17 16:57:34| Starting Squid Cache version 2.6.DEVEL-CVS for x86_64-unkno
wn-linux-gnu...
2006/05/17 16:57:34| Process ID 23052
2006/05/17 16:57:34| With 1024 file descriptors available
2006/05/17 16:57:34| Performing DNS Tests...
2006/05/17 16:57:34| Successful DNS name lookup tests...
2006/05/17 16:57:34| DNS Socket created at 0.0.0.0, port 32983, FD 8
2006/05/17 16:57:34| Adding nameserver 192.168.0.5 from /etc/resolv.conf

Under gdb I'm seeing this:

Program received signal SIGABRT, Aborted.
[Switching to Thread 47026893756432 (LWP 22594)]
0x00002ac54cbd75c5 in raise () from /lib64/libc.so.6
(gdb) bt full
#0 0x00002ac54cbd75c5 in raise () from /lib64/libc.so.6
No symbol table info available.
#1 0x00002ac54cbd8f10 in abort () from /lib64/libc.so.6
No symbol table info available.
#2 0x000055555557f4fb in xassert (msg=0x5842 <Address 0x5842 out of bounds>,
     file=0x5842 <Address 0x5842 out of bounds>, line=6) at debug.c:415
No locals.
#3 0x000055555557ca32 in comm_close (fd=24) at comm.c:647
         F = (fde *) 0x2ac54d61abd0
#4 0x00005555555c1baf in wccp2ConnectionShutdown () at wccp2.c:725
No locals.
#5 0x00005555555c1c05 in wccp2ConnectionClose () at wccp2.c:744
         service_list_ptr = Variable "service_list_ptr" is not available.
(gdb) frame 1
#1 0x00002ac54cbd8f10 in abort () from /lib64/libc.so.6
(gdb) frame 2
#2 0x000055555557f4fb in xassert (msg=0x5842 <Address 0x5842 out of bounds>,
     file=0x5842 <Address 0x5842 out of bounds>, line=6) at debug.c:415
415 abort();
(gdb) frame 3
#3 0x000055555557ca32 in comm_close (fd=24) at comm.c:647
647 assert(F->flags.open);
(gdb) frame 4
#4 0x00005555555c1baf in wccp2ConnectionShutdown () at wccp2.c:725
725 comm_close(theInWccp2Connection);
(gdb) frame 5
#5 0x00005555555c1c05 in wccp2ConnectionClose () at wccp2.c:744
744 wccp2ConnectionShutdown();
(gdb) info locals
service_list_ptr = Variable "service_list_ptr" is not available.
(gdb)

My WCCPv2 doesn't seem to work at all actually, config looks like this:

wccp2_router 192.168.0.1
wccp2_version 4
wccp2_forwarding_method 1
wccp2_return_method 1
wccp2_service standard 0
wccp2_outgoing_address 192.168.0.5

Seems that the server doesn't even attempt to talk to the router at all for some
reason :(

reuben
Received on Tue May 16 2006 - 23:48:48 MDT

This archive was generated by hypermail pre-2.1.9 : Thu Jun 01 2006 - 12:00:04 MDT