We need to turn off SNMP at our Solaris 10 (09/10) two-node Sun
cluster servers due to weak community string due to security
compliance. There are 2 local zones in each node. All together there
are 3 snmpd processes, i.e., one from the global zone/primary node and
two from the local zones.
# ps -ef | grep snmpd
root 8374 1 0 01:05:14 ? 0:01 /usr/sfw/sbin/
snmpd <---- from global / main cluster node
root 8377 1 0 01:05:14 ? 0:01 /usr/sfw/sbin/
snmpd <---- from local zone A
root 10884 1 0 01:08:44 ? 0:01 /usr/sfw/sbin/
snmpd <---- from local zone B
We tried to kill all snmpd processes by using "# kill -9 8374 8377
10884" at the global zones or by using "# kill -9 8377" / "# kill -9
10884" at each local zone. Unfortunately, none of the snmpd process
is killed and each process is "re-spawned" again with different
process number. Bummer !! Could anyone let us know how to kill or
turn off all snmpd processes above at our Solaris 10 server? If not,
what is the best way to remediate the weak community string in SNMP ?
Any workaround to kill all snmpd processes by means of updating the
files, e.g., /etc/sma/snmp/snmpd.conf or /etc/snmp/con/snmpd.conf, is
it do-able ?
The "snmpd" service has been completed disabled at all times.
disabled Mar_13 svc:/application/management/snmpdx:default
We appreciate in advance for your assistance.
Thanks,
Bill
You are seeing the sfw snmpd, not snmpdx.
man snmpd will give you the details, see the notes at the end.
--
Ian Collins
search for snmp services:
# svcs -a | grep -i snmp
> search for snmp services:
> # svcs -a | grep -i snmp
That won't find them, rather annoyingly. Searching for "management"
will find them, and some other things. Ian's suggestion about the man
page was spot on.
--
Chris