I had asked about an error I was getting about my token ring software
(see included message below).
Several people pointed out that this was a documented error in Suns
"Important Product Information" and it did not cause any problems.
It appears that the first error is caused by a bug in IP in which an
ethernet-encapsulated packet (versus token-ring encapsulated) is sent,
and the second error is caused by IP attempting to use ethernet-type multicasts
on the token ring which token ring does not support so it's just ignored.
I haven't figured out why the machine slowly dies when a few users
run wabi on it yet. It's a sparcserver 20 with 128mb ram and 2 processors.
I did find a parameter called tr_mbufs which I have now
increased to the recommended 500 from 64, and hopefully that will do the trick.
Thanks to:
Darren.Kinley@BC.net (J. Darren Kinley)
Scott Spevacek <spev@badger.state.wi.us>
Henry Katz <hkatz@panix.com>
jmd@san-jose-1.firstimage.com (Joe Doyle)
Jon.Masyga@seag.fingerhut.com (Jon Masyga)
jerry.springer@valcom.com (Jerry Springer)
Al.Venz@seag.fingerhut.com (Al Venz)
Casper Dik <casper@Holland.Sun.COM>
etolido@cns-ss10.cnsinc.com (Erik J. Tolido)
----- Begin Included Message -----
> From sun-managers-relay@ra.mcs.anl.gov Wed Aug 23 10:09 PDT 1995
> Date: Tue, 22 Aug 95 18:07 GMT+4628841:32
> From: heather@metaxa.wimsey.com (Heather Scott)
> To: sun-managers@eecs.nwu.edu
> Subject: TRI/S error message
>
>
> Hello out there!
>
> I have a network of Sunos 4.1.3 machines, and am adding some Solaris 2.4
> machines to it.
>
> I am having a problem with the token ring software TRI/S 3.0.2 on the
> Solaris 2.4 boxes.
>
> I realize that most of you are luckier than I and don't have to use token ring.
>
> I get this message at boot up. Has anyone seen it before, know what it means,
> or how to fix it?
>
> Aug 21 16:00:59 pop unix: tr0 at sbus0: SBus slot 1 0xc00000 and SBus slot 1 0x0 and SBus slot 1 0x400000 SBus level 4 sparc ipl 7
> Aug 21 16:00:59 pop unix: tr0 is /iommu@f,e0000000/sbus@f,e0001000/tr@1,c00000
> Aug 21 16:00:59 pop unix: ip_rput: DL_ERROR_ACK for 29, errno 1, unix 0
> Aug 21 16:00:59 pop unix: ip: joining multicasts failed on tr0 - will use link layer
> Aug 21 16:00:59 pop unix: broadcasts for multicast
>
> I have installed patch #102463-01 which is supposed to fix some tr problems, but
> I still get the error.
>
> The machine works ok when booted and can talk to the network, but I suspect there
> are network problems, because when several people try to run Wabi on this
> machine, the error "nfs server crunch not responding" appears and it never
> reconnects.
>
> Thanks!
>
>
> --
>
> ****************************************************************************
> * Heather Scott, SysAdmin, *
> * Corporate Actuarial, ICBC, Vancouver, B.C. *
> ****************************************************************************
>
----- End Included Message -----
--**************************************************************************** * Heather Scott, SysAdmin, * * Corporate Actuarial, ICBC, Vancouver, B.C. * ****************************************************************************
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:10:32 CDT