Hi,
We seem to experience a problem when we remove a yp_slave/bind_slave server
from the network. Currently we have a master ypserver, and one yp_slave
server (which is this particular machine), a master bind server and a number
of bind slave servers. All our systems are DU4.0d or e, with a couple of
Solaris 2.6's. However there is one particular machine (a DEC 1000a) with
DU4.0 which only handles a number of nfs connections. There have been a
couple of times when we have had to shut down the yp_slave/bind_slave
server. Both times the network has ground to a halt. My impression is its
the software on the DU4.0 machine not being able to cope with the fact that
the yp_slave/bind_slave has gone down and swap to an alternate server, at
which time all its nfs connections grind to a halt, and this impacts on the
rest of the network that has an nfs connection to that machine.
Would someone like to comment on what they think might be happening.
Dennis
------------------------------------------------
Dennis Macdonell |
Systems Administrator |
AUSLIG | "Bad is never
em: mcdonell_at_auslig.gov.au | good until worse
ph: 02 6201 4326 | happens"
fax: 02 6201 4377 |
------------------------------------------------
Received on Fri Jul 16 1999 - 06:52:47 NZST