No solution, but more data. We have an asterisk server at a branch office (where it will reside at the end of the evaluation) and the NTP server at the main office. The NTP server runs under freeBSD. DHCP info at the main office is offered up by a Windows 2000 server and at the branch office by a cisco PIX 501. A VPN connection is used between the offices. The main office network is a hive of activity, the branch office network is relatively quiet. The two GXP-2000 units running 1.0.1.9 produced the same repeatable results (~ 10 power cycles for each configuration) as follows:
1) Static number on the main office network = clock wrong, registered with asterisk 2) DHCP number on the main office network = clock wrong, failed to register with asterisk 3) Static number on the branch office network = clock correct, registered with asterisk 4) DHCP number on the branch office network = clock correct, registered with asterisk A final test was to configure the phone for DHCP on the main office network, unplug it, take it to the branch office, plug it in = results as above. Taking it back to the office resulted in the same failure mode as above. Since the phone was set to DHCP there was no need to modify anything. Since this happens with 2 phones and is repeatable it tells me it is with the phone. For purposes of our application the phones work since they will be at the branch office, however knowing the bug is there is reason for concern. Perhaps there are other packets on our main office network which are causing stack problems with the phone. A polycom 601 in the same 4 scenarios worked perfectly (but they are more than twice the price). Tony. >On Dec 31, 2005, at 7:28 AM, Ross C wrote: >> Peter, >> >> After upgrading to 1.0.1.13 I had some miscellaneous problems on >> one of my >> GXP-2000's--it would grab an IP address, but it wouldn't get the >> time/date, >> it wouldn't register, blah blah blah. I could access the web >> interface OK, >> so it wasn't a network issue (I don't think). Anyway...I ended up >> resetting >> to factory defaults and all is well now. Maybe try that? That has >> solved >> some other problems I've had as well. >I just got a 2000 which does exactly this (our first for evaluation.. >which is somewhat disappointing thus far). I could see in a packet >sniffer a weird cycle of DHCP requests like it got an IP but kept >retrying? A power cycle doesn't solve the problem (it's had many, >and dozens of software resets). A reset with the MAC input doesn't >work either for me. The phone was at an older FW when I got it >(ending in .9, I think) and then updated to to the latest stable (.12 >I think off the top of my head). Btw- the firmware update was a >pain. HTTP updates were hitting the server (Apache) with 'bad >request' results. I needed to set up my own tfpt server to make it >work. Off lan updates weren't working, either, in any case. >The phone will register and work when it has a static address >assigned, but not when set for DHCP. In all cases, the clock is >always wrong. I can see with a packet sniffer that the NTP request >is sent and received, but with no effect on the phone display. >Was there a resolution to this issue? The GXP-2000 seems to be a >very popular phone, so I can't imagine others on the list not >experiencing this? Or is this part of a batch with unresolvable >problems that I need to send back to the seller? >Thanks! TGIF! :') > >Phil _______________________________________________ --Bandwidth and Colocation provided by Easynews.com -- Asterisk-Users mailing list To UNSUBSCRIBE or update options visit: http://lists.digium.com/mailman/listinfo/asterisk-users
