sigurasg@rhi.hi.is (Sigurdur Asgeirsson) (05/14/91)
We have an fx running A/UX 2.00 at my university, it's connected via ethernet
to the other unix machines at campus, and via localtalk to an AppleShare server
and an old laserwriter. In the past 4-5 weeks the appletalk interface has gone
down at least 3 times. What happens is that all of a sudden it won't print to
the LW or talk to the AS server, and when I try to find out what's happening
with appletalk -s, it aborts with the message "no more processes" or something
to that effect. Appletalk -d does the same thing and atlookup sees nothing on
the net. So far the only solution has been to reboot the machine.
Does anyone know why this happens, and how to:
a: prevent it from happening, or
b: bring the appletalk interface up w/o rebooting
--
sigurasg@rhi.hi.is | "Well you know, C isn't that hard, void (*(*f[])())() for
Sigurdur Asgeirsson | instance defines f as an array of unspecified size, of
| pointers to functions returning a pointer to a function
| returnig void... I think"liam@dcs.qmw.ac.uk (William Roberts) (05/15/91)
In <3141@krafla.rhi.hi.is> sigurasg@rhi.hi.is (Sigurdur Asgeirsson) writes: >In the past 4-5 weeks the appletalk interface has gone >down at least 3 times. What happens is that all of a sudden it won't print to >the LW or talk to the AS server, and when I try to find out what's happening >with appletalk -s, it aborts with the message "no more processes" or something >to that effect. Appletalk -d does the same thing "no more processes" corresponds to EAGAIN, which means that there is some supposedly temporary resource limit: it gets my prize for the most unhelpful error message produced by A/UX (adb translates this into "Try again" which is possible worse!). Unless someone from Apple would care to answer this question for a change, I'd say that there is nothing you can do but kill a lot of processes and probably that won't help either. There are NO USEFUL DIAGNOSTICS available to tell you what resource has run out, and nothing appears on the console. None of the programs like ps, pstat, sar etc can tell you anything helpful (even if you do manage to run one without getting responses like "No more processes" or just plain "Killed", which is how the shell handles fork failures). Specific questions are: 1) Am I right about the diagnostics, or is there something else we should try? 2) What resource is running out? 3) Does 2.0.1 have kernel changes which fix this in some way (in which case the answer to 2 should be easy to find)? P.S. Could someone have Lonny Doi read his mailbox, please? -- William Roberts Internet: liam@dcs.qmw.ac.uk Queen Mary & Westfield College UUCP: liam@qmw-dcs.UUCP Mile End Road AppleLink: UK0087 LONDON, E1 4NS, UK Tel: +44 71-975 5234 (Fax: +44 81-980 6533)
andrew@jhereg.osa.com (Andrew C. Esh) (05/18/91)
In article <3141@krafla.rhi.hi.is> sigurasg@rhi.hi.is (Sigurdur Asgeirsson) writes: > > We have an fx running A/UX 2.00 at my university, it's connected via ethernet >to the other unix machines at campus, and via localtalk to an AppleShare server >and an old laserwriter. In the past 4-5 weeks the appletalk interface has gone >down at least 3 times. What happens is that all of a sudden it won't print to >the LW or talk to the AS server, and when I try to find out what's happening >with appletalk -s, it aborts with the message "no more processes" or something >to that effect. Appletalk -d does the same thing and atlookup sees nothing on >the net. So far the only solution has been to reboot the machine. > Does anyone know why this happens, and how to: > a: prevent it from happening, or > b: bring the appletalk interface up w/o rebooting > > >-- >sigurasg@rhi.hi.is | "Well you know, C isn't that hard, void (*(*f[])())() for >Sigurdur Asgeirsson | instance defines f as an array of unspecified size, of > | pointers to functions returning a pointer to a function > | returnig void... I think" I'd be willing to bet that the Ethernet card in it is from Racal/Interlan. We had one in here doing pretty much the same thing. I am told by a salesman that the Farallon ethernet card has been tested with A/UX, but I haven't personally had the pleasure yet. -- Andrew C. Esh andrew@osa.com Open Systems Architects, Inc. Mpls, MN 55416-1528 Punch down, turn around, do a little crimpin' (612) 525-0000 Punch down, turn around, plug it in and go ...