deepak_garg@rediffmail.com (deepak) writes:
> This process is getting repeated over and over again.
> I feel my USB TA is not understanding the option 0x17 (Link
> discriminator for BACP ...uh???).

Because it's mangled. 0x17 01 91 isn't a possible legal LCP option.
The length octet (01) isn't at least the minimum legal value (02).

Sounds to me like a bug in the equipment this ISP is using.

--
James Carlson, IP Systems Group
Sun Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677

deepak_garg@rediffmail.com (deepak) writes:
> 00:38:35:76 39 bytes B1 PPP REMOTE LCP Config-Req Id = 0x72
> Option(1)..Maximum-Receive-Unit:1600
> Option(5)..Magic Number:0x33B5BC55
> Option(17)..Multilink-MRRU:1600
> Option()..Type:0x17 01 91
> Option(3)..Authentication:PAP
> Option(19)..MP-EP-Discr:Class=3 08 00 03 04
>
> 00:38:35:76 39 bytes B1 PPP LOCAL LCP Config-Rej Id = 0x72
> Option()..Type:0x17 01 91

Ah, wait. I think I misinterpreted that strange debug output. What
the TA really received was (hex) 17 02 01 91 -- option 0x17 with data
01 91. That's legal. My mistake.

> This process is getting repeated over and over again.
> I feel my USB TA is not understanding the option 0x17 (Link
> discriminator for BACP ...uh???).

This still points to a likely ISP bug. Implementations are _required_
to accept and respond to Configure-Reject messages.

--
James Carlson, IP Systems Group
Sun Microsystems / 1 Network Drive 71.234W Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757 42.497N Fax +1 781 442 1677