Appletalk service getting knocked out

[Anonymous]

Weird problem with this one. we have a dual G4 with a gig of mem,running 10.28 server. we a firewire 800 card to run to lacie firewire 500gb drives. for some storage and file serving. every since then it randomly knocks out the appletalk service , knocking off all the mac clients. Does not effect the windows sharing. I can restart appletalk and it is fine for awhile. but this is causing a huge problem because files are not getting saved . any ideas?

[Anonymous]

were using appletalk to communicate with 40 some os x machine’s for file serving and storage. it has one NIC card.
thanks for any help

[richardab1967]

This sounds very similar to problems we are encountering with 10.3.5 server on G5 tower. Randomly, sometimes weekly, sometimes daily we are all (12 clients 10.3.3-10.3.5) disconnected from the server, the server does not crash but AFP has restarted.

Graphs in server admin show quite a lot of cpu usage and network traffic around the time of disconnect although nothing especially intensive is being done.

Also afp connections shows two connections per person, the duplicated entry has same name, type, address and connected time, but idle time is often 0.00 for one entry and a figure for the other. What is all this about?

The AFP crash logs always reports:

Host Name: server.winch.dea
Date/Time: 2005-01-14 10:29:36 +0000
OS Version: 10.3.5 (Build 7M34)
Report Version: 2

Command: AppleFileServer
Path: /usr/sbin/AppleFileServer
Version: ??? (???)
PID: 15264
Thread: 0

Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_INVALID_ADDRESS (0x0001) at 0x0109e004

Thread 0 Crashed:
0 AppleFileServer 0x000a7e6c 0x1000 + 0xa6e6c
1 AppleFileServer 0x0003f58c 0x1000 + 0x3e58c
2 AppleFileServer 0x00053eb0 0x1000 + 0x52eb0
3 AppleFileServer 0x00053fa0 0x1000 + 0x52fa0
4 AppleFileServer 0x0005153c 0x1000 + 0x5053c
5 AppleFileServer 0x00006828 0x1000 + 0x5828
6 AppleFileServer 0x0000669c 0x1000 + 0x569c

I have no knowledge of logs so I was hoping someone out there can translate this.

Would be grateful for any help as we are using network home directories which means everyone not only loses connection to their files, their desktops go down and require a force restart.

[Anonymous]

I note that the messages are 4 years apart. Any solutions for this? For me it’s at least once a day, not exactly the same time, but frequently during the same “period” (we are a school.)

DP G4 -533Mhz. Gigabit Ethernet w/extra 4port PCI ethernet card. One drive 200Gb SATA, other drive RAID mirror= 2 SCSI 40gb. 40+ macs w/ OS X Server 10.2.8. 1.5Gb RAM.

thanks in advance! mailto:[email protected]

[cyngus]

While I can’t provide you a solution, from the crash log its basically saying that the AFP server process tried to access a memory address outside of its allowed range. Internally the process is probably improperly dereferencing a pointer or calculating an offset from a pointer improperly.

[hhavel]

I manage 5 different schools. At one school we had similar problems with AFP disconnects One of the problems we found was some of the switches and or routers had the spanning tree protocol enabled. According to Cisco’s website “Spanning-Tree Protocol is a link management protocol that provides path redundancy while preventing undesirable loops in the network. For an Ethernet network to function properly, only one active path can exist between two stations. ” As soon as we disabled this protocol almost all disconnects stopped. I hope this helps someone. I was pulling my hair out, checking server settings, and the whole time I had nothing to do with it.

[Author]

Posts