Red Hat Bugzilla – Bug 7393
CCP Compression fails; connection becomes useless
Last modified: 2008-05-01 11:37:53 EDT
Since upgrading to 6.1 I have been having massive trouble with CCP compression. It took a while to figure out what the problem was, but it appears to be CCP... The problem goes away when I add noccp to /etc/ppp/options, and returns when I reenable it. I have two different dial-in accounts (ISP & work), and both have the same problem. I know that the RAS servers are different, so the problem must be my installation. The connection will work fine for, oh, saw five minutes or so depending on how heavy I'm hitting it. Then it'll just "hang". tcpdump shows that there's still data being exchanged, but I can't do anything (web, mail, news, etc). Once the connection hangs PPP starts spewing the following debug messages: Nov 28 11:33:12 lancelot pppd[3425]: sent [CCP ResetReq id=0x5] Nov 28 11:33:12 lancelot pppd[3425]: rcvd [Compressed data] 04 7f 0b 40 f2 98 0a cf ... Nov 28 11:33:12 lancelot pppd[3425]: rcvd [CCP ResetAck id=0x5] Nov 28 11:33:16 lancelot pppd[3425]: rcvd [Compressed data] 00 00 17 91 40 00 56 e2 ... Nov 28 11:33:16 lancelot pppd[3425]: sent [CCP ResetReq id=0x6] Nov 28 11:33:16 lancelot pppd[3425]: rcvd [Compressed data] 00 01 0b 40 f2 98 0a b9 ... Nov 28 11:33:16 lancelot pppd[3425]: rcvd [CCP ResetAck id=0x6] Nov 28 11:33:25 lancelot pppd[3425]: rcvd [Compressed data] 00 00 17 91 40 00 56 e3 ... Nov 28 11:33:25 lancelot pppd[3425]: sent [CCP ResetReq id=0x7] Nov 28 11:33:25 lancelot pppd[3425]: rcvd [Compressed data] 00 01 0b 40 f2 98 0a b9 ... Nov 28 11:33:25 lancelot pppd[3425]: rcvd [CCP ResetAck id=0x7] Nov 28 11:33:41 lancelot pppd[3425]: rcvd [Compressed data] 00 00 17 91 40 00 56 e0 ... Nov 28 11:33:41 lancelot pppd[3425]: sent [CCP ResetReq id=0x8] Nov 28 11:33:41 lancelot pppd[3425]: rcvd [Compressed data] 00 01 0b 40 f2 98 0a b9 ... Nov 28 11:33:41 lancelot pppd[3425]: rcvd [CCP ResetAck id=0x8] Pertinent info: ppp-2.3.10-3 initscripts-4.63-1 I normally compile my own kernel, but haven't done so on this box yet, so I'm running the stock 6.1 kernel (2.2.12-20). I can limp along with noccp, but this makes a slow 56k connection even more painful!
please close this "bug"... i complained to my phone service provider and they detected a problem on the line... a technician came out and did something (hey, i was at work ;-D) they must have corrected my problem. i re-enabled ccp and things seem to be working normally. sorry.... i trouble-shot everything i had control over... <sheepish grin>