danny@garnet.berkeley.edu (Danny "Iceman" Lee) (02/02/89)
Hi there !!! I am experiencing problems on an Appleshare network. I am currently using Appleshare 2.0.1 with System 6.02 on 15 MacPluses, 4 SE's and 2 Mac II's hooked up via Appletalk and Phonet. The problem is that some machines can not recognize the laserwriters while others can. The same problem occurs with some machines "logging" onto the server; some Pluses see it while others don't. I've checked all the connections on each machine and they seem to be fine. I've also tried reinstalling the laserwriter drivers as well as the Appleshare drivers but no dice. Any help would be greatly appreciated. Thanx in advance. Please email me directly if possible. danny@garnet.berkeley.edu
denbeste@bgsuvax.UUCP (William C. DenBesten) (02/03/89)
From article <19807@agate.BERKELEY.EDU>, by danny@garnet.berkeley.edu (Danny "Iceman" Lee): > I am experiencing problems on an Appleshare > network. [...] The problem is that some machines can not recognize the > laserwriters while others can. The same problem occurs with some > machines "logging" onto the server; some Pluses see it while others > don't. There are 2 causes of this. First, check to make sure that the everything is properly connected electrically. Can the two machines on each end see each other or a common point, such as a LaserWriter or file server. Can the machine at the end of each branch see this common point, or a point that can see the common point. Keep in mind that if it looks physically connected, it does not necessarily mean that it is electrically connected. Also make sure that there are not any shorts. Some communication can go on, even in the presence of a short. I suspect that communication can not cross the short, however, When I set up our network, I disconnected all the nodes and made sure that the resistance between the two wires was infinite. I then shorted the wires at one node and made sure that everyone had low resistance. Make sure that you remove the shorting block when you are done. The second thing to check for is reflections on the network. Do you have any branches without a phone-net box on the end of them. Are all splices soldered? When you install a branch, do not cut the backbone, just strip the wire. Is there a terminating resistor on each end of your network? Did you use the same guage wire everywhere. Every splice, particularly between wires of different guages is a bad thing. Install a terminating resistor on a machine that can not see the server. Does this change what it can see? Does this change what others can see? If it does not solve all the problems, note how the resistor changed the network and go onto the next machine. After you have tried the troubled machines, try the untroubled machines. If you don't find the one perfect spot, put one in the spot that does the most good, and hunt down a second spot to solve the rest of the problems. In my case one resistor on the back of one machine solved all the problems. My network is set up as a 22 gauge wire backbone in the hallway. Each office has a 30 foot modular wire from the mac to a modular jack attached via screws to the uncut backbone in the hallway. Next time, I would run the backbone into each office and make the branches 4 feet long, to minimize the different wire gauges. I could also instruct people to disconnect from the wall when they remove their machine. I have found that Mac IIs are less succeptable to problems than pluses or SEs. Moving a II (if you have one) around with you may help determine if the network is physically connected. I would also recommend running responder (free with system 6.0) on each machine, and using interpoll (published by apple, about $100) to determine which machines can communicate with which machines, and if the packets can travel without getting lost or corrupted. Once, after I performed major surgery on our network, I found that to restore order, I had to shut off all the macs and restart them before all the problems went away. Perhaps this would be the best starting point. I suspect that the problem was that connecting two the halves caused two machines to share one address. I have figured out most of this by trial and error. Some of it came from the little book that farallon gives away with their connectors. -- William C. DenBesten | denbeste@bgsu.edu Dept of Computer Science | CSNET denbeste%andy.bgsu.edu@relay.cs.net Bowling Green State University | UUCP ...!osu-cis!bgsuvax!denbeste Bowling Green, OH 43403-0214 |