10.3.9 Keeps crashing, only on specific machines

[valid377]

Im running a small network of roughly 8 computers, all running OSX 10.3.9.
3 of them keep crashing constantly, the window dims and tells me to restart.
Ive repaired permissions, defragged, and even scanned for viruses and still no help, they still keep crashing.

One thing they all do have in common is this:

Mar 8 15:37:07 localhost kernel: NHCI: setup TCP connection port 3842 IP 192.168.1.100
Mar 8 15:37:07 localhost kernel: NHCI Status[192.168.1.100]: nbsInitial
Mar 8 15:37:07 localhost kernel: Starting TCP session with server, client[machine1]
Mar 8 15:37:07 localhost kernel: KNUdoClientReceive: receive(len) returned 36
Mar 8 15:37:07 localhost kernel: KNUdoClientReceive done
Mar 8 15:37:12 localhost kernel: NHCI: setup TCP connection port 3842 IP 192.168.1.100
Mar 8 15:37:12 localhost kernel: NHCI Status[192.168.1.100]: nbsInitial
Mar 8 15:37:12 localhost kernel: Starting TCP session with server, client[machine1]
Mar 8 15:37:12 localhost kernel: KNUdoClientReceive: receive(len) returned 36
Mar 8 15:37:12 localhost kernel: KNUdoClientReceive done
Mar 8 15:37:17 localhost kernel: NHCI: setup TCP connection port 3842 IP 192.168.1.100
Mar 8 15:37:17 localhost kernel: NHCI Status[192.168.1.100]: nbsInitial
Mar 8 15:37:17 localhost kernel: Starting TCP session with server, client[machine1]
Mar 8 15:37:17 localhost kernel: KNUdoClientReceive: receive(len) returned 36
Mar 8 15:37:17 localhost kernel: KNUdoClientReceive done
Mar 8 15:37:22 localhost kernel: NHCI: setup TCP connection port 3842 IP 192.168.1.100
Mar 8 15:37:22 localhost kernel: NHCI Status[192.168.1.100]: nbsInitial
Mar 8 15:37:22 localhost kernel: Starting TCP session with server, client[machine1]
Mar 8 15:37:22 localhost kernel: KNUdoClientReceive: receive(len) returned 36
Mar 8 15:37:22 localhost kernel: KNUdoClientReceive done
Mar 8 15:37:27 localhost kernel: NHCI: setup TCP connection port 3842 IP 192.168.1.100
Mar 8 15:37:27 localhost kernel: NHCI Status[192.168.1.100]: nbsInitial
Mar 8 15:37:27 localhost kernel: Starting TCP session with server, client[machine1]
Mar 8 15:37:27 localhost kernel: KNUdoClientReceive: receive(len) returned 36
Mar 8 15:37:27 localhost kernel: KNUdoClientReceive done


They all seem to communicate to another machine with the IP address of 192.168.1.100. I dont have any machines under that IP address. I am not using DHCP on the router and the router is 192.168.1.1. I do not fully understand why these 3 machines are spamming these logs in Console. All my other machines do not show these in the Console system logs.

Ive searched in the forums for something even remotely similar. Nothing came up. If I disconnect the network cable it stops and gives me the following:


Mar 8 15:45:36 localhost kernel: UniNEnet::monitorLinkStatus - Link is down.
Mar 8 15:45:36 localhost configd[101]: AppleTalk shutdown
Mar 8 15:45:36 localhost /usr/sbin/AppleFileServer: AFPServer::AbnormalATListenerShutDown - clear listener
Mar 8 15:45:36 localhost configd[101]: AppleTalk shutdown failed, status = 71 (retrying)
Mar 8 15:45:37 localhost configd[101]: AppleTalk shutdown
Mar 8 15:45:37 localhost configd[101]: AppleTalk shutdown complete
Mar 8 15:45:40 localhost configd[101]: posting notification com.apple.system.config.network_change
Mar 8 15:45:40 localhost mach_init[2]: Server 0 in bootstrap d03 uid 0: "/usr/sbin/lookupd": exited as a result of signal 1 [pid 209]
Mar 8 15:45:40 localhost configd[101]: executing /System/Library/SystemConfiguration/Kicker.bundle/Contents/Resources/set-hostname
Mar 8 15:45:40 localhost lookupd[470]: lookupd (version 324.13) starting - Wed Mar 8 15:45:40 2006
Mar 8 15:45:41 localhost set-hostname[475]: setting hostname to machine1.local

When i reconnect the network cable, it begins spamming the messages again and eventually the machine crashes. Can anyone help?

 

[valid377]

Just to let everyone know, I've looked everywhere for a similar problem such as this and I never found one, so this may be the first of its kind. <shrug>

The machines that had been spamming packet sends/requests to an IP address all crash at the same time everyday. One crashes at exactly 3:04pm, another crashes at exactly 4:25pm everyday. I shut down everything, even with extensions off. It still spams as long as its plugged to a network.

I even changed my router address, this only changed the packet sends/request to an IP address of 0.0.0.0 and upped it from 31 to 61. Ive repaired permissions, defragged, checked for viruses and it still crashes everyday at the same time like clockwork.

THE SOLUTION: reinstall OSX.

All machines were running OSX10.3.9. I reinstalled 10.3 and kept the Previous System Folder just in case. I then upgraded to 10.3.9. It fixed it, and hasnt crashed since. Must be some buggy service or maybe a corrupted library file. Who knows. This was the only solution I found to fix it.

 

[Cryptonomicon]

Howdy,

I just started experiencing this same exact problem on one of my machines over the past couple of days. I am betting that you have a Keyspan USB Server on your network (or did at some point), and have the software drivers installed for it on your Macs. Port 3842 is used by the USB Server for communications.

I just changed the IP address of my USB Server recently - around the time the crashes started. Furthermore, the IP address that appears for me in my logs (which points to nothing now) used to be the address of the USB Server.

The solution was to go into /Library/Preferences/com.keyspan.KeyspanNHCI.plist and remove the <key>server access</key> and following lines between <array> and </array>. You should note that the incorrect IP is in the array block under "accessURL".

Reinstalling Mac OS X fixes things because you blow away the Keyspan USB Driver and its preferences.