|
After futher testing, I found that this is currently not supported.
I would like to make a suggestion to engineering to allow this type
of test in the future. Allowing alarming of objects that have
process ids can identify if a server process is currently available
for a varity of client-server software that Digital offers.
Also I have had several customer requests during demonstrations
for a feature like this.
Please review the enclosed output from me alarm definition and
let me know if my results are correct.
...dk
$ set ver
$ manage /enterprise
DECmcc (V1.1.0)
create mcc 0 alarms rule poncho_pcfs_state -
expression = (CHANGE_OF(NODE4 PONCHO OBJECT PCFS PROCESS ID , -
*, *), AT EVERY 00:02) , -
procedure = mcc_common:mcc_alarms_mail_alarm.com , -
exception handler = mcc_common:mcc_alarms_mail_exception , -
parameter = mollyb::system , -
category = "Pathworks for MSDOS" , -
description = "Pathworks for MSDOS Server Services not available" , -
queue = mollyb_batch , -
perceived severity = CRITICAL , -
in domain = .act_epub
MCC 0 ALARMS RULE poncho_pcfs_state
AT 12-APR-1991 16:22:16
Data type of the attribute in the alarm expression is currently unsupported.
exit
$ exit
|
|
Are there any plans to allow creation of an alarm with an
expression checking to see if the PID of a Phase IV object
exists or has changed.
As stated in the above notes, Digital sells client-server
software. Many (and I mean several) clients have a home
office monitoring many remote sites. Each of these sites have
Pathworks for DOS running on a server. The home site that would
be running MCC, is not interested is seeing if a PC is on the
network. They are interested in knowing if LAD service at that
site is available to the network. One method of checking this
is to have an MCC alarm which would check to see if the PCFS
object has a PID associated with it on the remote server.
Currently this is not supported (.-1). Since Ditigal is strongly
pursuing Novell in network management, this seems like a defacto
feature.
Any comments...?
...dk
|