Yes, it does clarify things quite a bit! Going back to the USR TC
comparison, this might account for the relativly high Lost Carriers it sees
compared to the relativly low amount of Lost Carriers I see. Again,
looking at my disconnect stats on completed calls (i.e. that make it to a
Stop record in RADIUS), I see the following on a PM3 (with known SLIP users
exluded)
normal 91.31% 3741/4097
"User Request - Call Circuit Closed" 7.27% 298/4097
"Lost Carrier" 0.73% 30/409
"User Error - PPP NCP Active to Request" 0.66% 27/4097
"User Error - LAPM negotiation timeout" 0.02% 1/4097
or about 8% for today on this particular PM3. On the USR TC, I see
grep Acct-Terminate-Cause detail | sort | uniq -c | sort -n
26 Acct-Terminate-Cause = Session-Timeout
147 Acct-Terminate-Cause = Lost-Carrier
1120 Acct-Terminate-Cause = User-Request
Or about 11%... Off the same CO, same switch, same telco PRI service, and
presumably, a similar mix of client modems. Too bad we cant see what the
other NAS vendors are doing. It would help tremendously to see some real
world figures from Cisco and Ascend to see if they fair any better. 8-10%
of calls terminating abnormally (be that the client or server fault) is
high enough to drive customers to cable modems :-(
---Mike
**********************************************************************
Mike Tancsa, Network Admin * mike@sentex.net
Sentex Communications Corp, * http://www.sentex.net/mike
Cambridge, Ontario * 01.519.651.3400
Canada *
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.
Searchable list archive: <URL:http://www.livingston.com/Tech/archive/>