D-Link Forums
The Graveyard - Products No Longer Supported => Routers / COVR => DIR-655 => Topic started by: nedhedrick on January 05, 2009, 01:05:24 PM
-
DIR-655, HW ver A3. Unit was just purchased with FW 1.11. I have now upgraded to 1.21, with SecureSpot.
It works fine until I attempt to specify an NTP server. All I do is check the "Enable NTP Server" box, and specify server "us.pool.ntp.org". After doing so, I save and reboot, and the unit goes into a "reboot loop", rebooting itself every 30 seconds or so. It doesn't allow enough time to even get in and change the option back!
Solution I found was to unplug the cable modem. This stopped the rebooting. Then I can log into the unit once again and disable the NTP server. After that, I can plug my cable modem back in and everything is fine once again.
This behavior did not happen with the shipped FW (1.11), or with the non-SecureSpot FW DIR655A4_FW121B08.bin.
I'd like to use NTP to keep the clock correct, but...
Anyone else seen this behavior? Or is it a new bug?
-
I don't think its a bug (bugs tend to be very consistent and generic). Try reflashing the firmware again, this should solve the issue.
PS: leave 'beta' firmwares alone a go for the final. Beta's are not called beta's for nothing.
-
DIR-655, HW ver A3. Unit was just purchased with FW 1.11. I have now upgraded to 1.21, with SecureSpot.
It works fine until I attempt to specify an NTP server. All I do is check the "Enable NTP Server" box, and specify server "us.pool.ntp.org". After doing so, I save and reboot, and the unit goes into a "reboot loop", rebooting itself every 30 seconds or so. It doesn't allow enough time to even get in and change the option back!
Anyone else seen this behavior? Or is it a new bug?
There might be a bug here -- it'll be hard to reproduce.
I saw something like this a while ago (mine was not as severe as yours, it wouldn't reboot the unit, but did frack up my WPA connections which rely on the clock for timers) and I attribute it to a bad server in the us.pool.ntp.org set (this resolves to multiple addresses) ... I changed to time.nist.gov (which resolves to one) and stopped having the problem.
Now get this -- since then, unbeknown to me, I've apparently switched back to us.pool.ntp.org (I've been there for a while now, must have been in my saved settings)! I haven't had the problem.
-
OK...latest results. I re-flashed the FW with the same 1.21 code. The DIR-655 operated normally. Then I once again enabled the NTP server option by checking the box and entering the NTP URL "us.pool.ntp.org". When I clicked "Save", the unit immediately rebooted with no warning, and then proceeded to go into its loop. This time I timed it -- it was rebooting approximately every 45 seconds.
Again, to stop it, I had to disconnect the cable modem. Only then I could login and turn off the NTP option. When I reconnected the cable modem, all was fine once again.
Seems to be consistent to me.
-
I restored to factory defaults and tried to reproduce this. No joy.
Firmware Version : 1.21, 2008/11/13 H/W A2
1. What result do you get from the command "nslookup us.pool.ntp.org"?
Look at your DNS servers on the Status - Device Info page ...
2. What result do you get from the command "nslookup us.pool.ntp.org (ISP Primary DNS Server)"?
3. What result do you get from the command "nslookup us.pool.ntp.org (ISP Secondary DNS Server)"?
example -- (the command would be the same in Windows or Linux and probably OSX)
$ nslookup us.pool.ntp.org 4.2.2.1
Server: 4.2.2.1
Address: 4.2.2.1#53
Non-authoritative answer:
Name: us.pool.ntp.org
Address: 75.144.70.35
Name: us.pool.ntp.org
Address: 192.157.38.60
Name: us.pool.ntp.org
Address: 65.255.217.202
Name: us.pool.ntp.org
Address: 66.250.45.2
Name: us.pool.ntp.org
Address: 72.167.54.201
All three outputs will be different, but they should all return multiple addresses...
-
What happens if you use D-Links NTP server?
-
The three NSLOOKUP commands yield:
1. "nslookup us.pool.ntp.org"
DNS request timed out.
timeout was 2 seconds.
Server: UnKnown
Address: 192.168.0.1
DNS request timed out.
timeout was 2 seconds.
DNS request timed out.
timeout was 2 seconds.
Non-authoritative answer:
Name: us.pool.ntp.org
Addresses: 208.75.88.4
64.202.112.75
65.255.217.202
66.246.72.94
75.144.70.35
2. "nslookup us.pool.ntp.org (ISP Primary DNS Server)"
Server: cdns2.cox.net
Address: 68.105.28.12
Non-authoritative answer:
Name: us.pool.ntp.org
Addresses: 192.157.38.60
217.160.252.91
64.73.32.134
72.167.54.201
74.53.76.34
3. "nslookup us.pool.ntp.org (ISP Secondary DNS Server)"
Server: cdns7.cox.net
Address: 68.105.29.12
Non-authoritative answer:
Name: us.pool.ntp.org
Addresses: 65.255.217.202
72.167.54.201
75.144.70.35
192.157.38.60
64.73.32.134
-
I tried both D-Link NTP servers (ntp1.dlink.com and ntp.dlink.com.tw), and the enabling function works fine. Based on the log entries, the updates are successful and the time is synchronized.
-
I also tried time.nist.gov, and it works fine. But when I use us.pool.ntp.org the system immediately reboots without any warning, and then reboots every 45 seconds or so. Seems to happen every time I try it.
-
Possibly a garbled response from the NTP host...
-
Packet capture is needed.
-
Funchords: I am using FW 1.21 dated 09/11/2008. This is the version downloaded from the D-Link website. Also I have HW level A3.
Lycan: Not sure how I could get a packet trace...this is a "simple" home environment and I don't have the diagnostic tools to go too deep!
-
ftp://ftp.dlink.com/Gateway/dir655/Firmware/dir655_firmware_121.zip (dated December 01 2008 on their website) is the 13 Nov 2008 version that I referred to.
-
Cannot reproduce the reboot sequence with the us.pool.ntp.org time server. Works perfectly OK...
And most of all, I can imagine more serious issues to be addressed by Dlink tech (no offense, nedhedrick. I can be too rude sometimes ;D). Just pick a different NTP server (http://support.ntp.org/bin/view/Servers/StratumTwoTimeServers).
-
just run ours.........
-
just run ours.........
Others are better (dunno why though) ;) LOL
-
I've circumvented the issue by specifying a different NTP site.
I can't believe it would be that consistent from us.pool.ntp.org, since different servers are involved. However, it certainly could be something in the DNS response from the COX servers assigned to me.
No offense taken...I did a stint in development for a time so I'm aware of "priorities". I'm just glad it can be circumvented easily.
Thanks for the help in isolating it!