I agree. [time(1)'s trimmed]
# via telnet
$ time pmwho-a-nas -a a-nas >/dev/null
0:01.52elapsed
# via a pmconsole command session
$ echo "show session" | time pm_command -nostop a-nas a-nas-password >/dev/null
0:01.91elapsed
# via a series of SNMP queries
$ time snmpwalk a-nas a-nas-community .1.3.6.1.4.1.307.3.2.1.1.1.4 >/dev/null
0:00.39elapsed
>that will log into all the PMs and stay logged in,
That might cause problems. PM's don't handle many telnet sessions
very well, i.e, a couple of debugging sessions and a couple of
monitors like this and it would be "full". I'd suggest using an SNMP
based system, unless this telnet thing did all the automated stuff you
needed done for that NAS.
>Does anyone have any input on how this type prog would scale,
You'd probably be limited mostly by memory and CPU on your monitor
system, that could severly limit scalability if the program didn't do
all the automated stuff itself. An SNMP based system would probably
scale better in general. Decentralizing can help a lot, i.e., a
monitor system in each HUB/POP.
-
To unsubscribe, email 'majordomo@livingston.com' with
'unsubscribe portmaster-users' in the body of the message.