I don't think it's the phone lines this time, I think it's some
software problem with UUCP on cbosgd. It started happening with
our most heavily used connection (seismo) 3 weeks ago, and then
last Saturday it started happening to topaz and several other
backbone neighbors (decvax, sdcrdcf, etc.) I thought at one time
it had something to do with sequence numbers overflowing (our
/usr/spool/news/.Sequence/cbosgd said "cfff") but the problem
persists after clearing out the files.
I can call out from cbatt, a 3B2/400, on the same phone lines and
it works fine. We are making an effort to move the UUCP load from
cbosgd to cbatt, but we had hoped for a more orderly progression
than this.
If you exchange mail regularly with cbosgd, I would appreciate it if
you would do the following. (1) Poll us every once in awhile (perhaps
once a day) until this problem can be worked out. (2) Put cbatt in
your L.sys or Systems file. The info is the same as for cbosgd, except
that the phone number is 3849, Datakit address cb/dk/cbatt. It currently
has only one 300/1200 baud modem, used for both dial-in and dial-out, but
we plan to move several of cbosgd's phone lines to it when we get cbatt
fully fixed up. (It still has some problems.) If you'd like to make
a new UUCP login on your system for cbatt, let me know what it should be.
cbosgd is running HDB UUCP, although it's sort of an old version. If
anyone recognizes the problem, I'd be very appreciative of any information
that would help track it down. Right now I'm stumped.
Mark
From the responses I got, all of the other sites that have experienced this
problem have Vadic modems. Does anybody have an idea in which procedure
this bug lies? It doesn't seem possible that it lies in aculib/vad.c.