SUMMARY: st jumbo patch reintroduced old problem

From: David Way (
Date: Tue Nov 26 1991 - 18:51:37 CST

Subject: SUMMARY: st jumbo patch reintroduces old problem

Sun Managers,

In article <> I wrote:

> Sun Managers,
> I just installed Sun patch # 100280-02, scsi-tape jumbo patch for 4.1.1
> on our 4/490. This was to correct a `psig action' panic which would
> occur during remote tape usage.
> Attempts to run remote dump now abort with:
> DUMP: Protocol to remote tape server botched (code Where are you??).
> rdump: Lost connection to remote host.
> DUMP: Bad return code from dump: 1
> According to the 100280-02 README, this problem was introduced by
> patch # 100268-02 (replacement binary for colldef(8)) and should have
> been fixed by 100280-02. My symptoms aren't precisely the same as
> those described, but closely similar (return code from dump is 1
> instead of 4). I wouldn't think that something like colldef(8)
> would affect the way dump/rdump works in any way, but I obtained
> 100268-02 anyway and installed it just for grins; as expected it had
> no effect - same result as above when running rdump.

Most of the responses I got suggested there that this problem was actually
being caused by the remote csh spawned by rdump and not rdump itself or the
patched st.o. This turned out to be true. Root's .cshrc on the host in
question contained a `biff n' line, and the problem went away when I removed
it. Apparently the shell gets confused when starting up as an rsh if
commands which attempt to locate the local stdout are executed, such as
biff, msg, stty. Makes sense to me ... In fact, this behavior is implied
in the BUGS section of the rsh(1c) manpage.

Thanks to those who responded:

simon@Aus.Sun.Com (Simon Woodhead) (Per Hedeland) (Phil Ritzenthaler) (Dan Lorenzini)
stern@sunne.East.Sun.Com (Hal Stern)

David Way  McDonald Observatory/Astronomy Dept.- Univ. of Texas, Austin 
(office) RLM 16.206 (voice) 471-7439 (internet)

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:16 CDT