Conference azur::mcc

3492.0. "some problems with DECmcc for ULTRIX V1.2 SSB" by MXOV06::LERIOS () Mon Aug 03 1992 23:11

Hi: 

I have been working with DECmcc for ULTRIX
V1.2 SSB and I have had some problems:

ALARMS:

The situation is the next one:

I want that more than one user receive the mail in the
moment that an alarm fired, if I write in the field ALARM
FIRED PARAMETERS only one mail address no problem, the
process send the mail to the user that I specified

but I can not do the  same if I specify more than one
mail address specification, by instance:

ALARM FIRED PARAMETERS= mxov09::JCrodiriguez mxov06::lerios

or

ALARM FIRED PARAMETERS= mxov09::JCrodiriguez,mxov06::lerios

I am sure that I have to modify the file:
/usr/mcc/mcc_system/mcc_alarms_mail_alarm.csh

but I hope someone can explain me a few more.....

2) If I start an alarm (comparison rule) using the commands interface,
   and without specify start time, I mean taking the default start time
   (right now), I receive a message telling me : 

    Normal Opertation Has Begun,

   However the time reported with this message is not the local time.

   Suppose that the local time is  Mon Aug 3 18:04:43 then

   the message I receive  tell me:

   the operation begun at Mon Aug 3 23:04:43
                           
   ah! I did this since a remote node, I mean,

   I entered to the system who has DECmcc running

   through a dlogin and created the alarm using the

   command interface.

When I open the Notification Services window (in the network management station)
I see the alarm event but with the same problem (start time problem)

ah !  if I ask by the status of the alarm (iconic map) since the system 

in which I am running DECmcc the status of it is disable !!!!

however every 15 minutes the alarm fired (looking the notification 

services window)    


RECORD:

I used the record process with one node of my configuration

starting before of this the background process specifying

the domain in which the selected node is included, the process worked

o.k.

when I tried to do the same but now

including about 30 devices , NODE4's,(in the same domain I used for
the previous record process) I received 3 messages:

a) Entity existence can not been determinated. In this point
   I askef for the status of all the devices and no problem  all of them
    were actived,

   in fact and using the comand:
   
   mcc>directory node4 *, in domain .domain.nafinsa

   all of the selected devices are registered (not partially)

b) Segment failure: no space left on device!!!
   
   space problem?

    what is the mean of this?

    using the df command I did not see space problem in
   
    my configuration.

3) SEMOP failure invalid argument....

ah! the process started o.k. for some of the devices !!!!


NOTIFY:

I created a notify request using a NODE4 of my configuration:

domain: .domain.nafinsa
entity list: node4 MYNODE
events: (counters zeroed, any events)

if I see the notify request window I found the state of such request

is enable, but nothing happened in my iconic map....

and using the command:

mcc>Display notify

I got a message telling me that there is not notify running.!!!

Any comments would be appreciated

thanks in advance and regards


Julio Lerios (digital Mexico)

I have forwarded this message to the supervisor responsible for the
alarms and notification modules.  Meanwhile:

>ALARM FIRED PARAMETERS= mxov09::JCrodiriguez mxov06::lerios

I haven't tried it, but perhaps 
   ALARM FIRED PARAMETERS= "mxov09::JCrodiriguez mxov06::lerios"
will work?


>b) Segment failure: no space left on device!!!

This is not a disk space problem, it is a semaphore problem.
There are instructions in the historian use guide describing how
to increase the number of semaphores available to the system.

>3) SEMOP failure invalid argument....

Same problem as 2b, just a different symptom.

Erik:

Thanks for your comments

I am going to test them...

regards

Julio Lerios

Ah

I forgot to mention the next situation:

Talking about record function.

the manual say that the format we have to uset to

specify time is the next one:

04-OCT-1992 09:00 in other words: dd-mmm-yyyy hh:mm:ss

the same format should be used for all the times we have to use:
start time,
polling time
end time.


ah! but this does not work.. I mean using this time specification

I can not start the record process

the format I have used is (at least in DECmcc for ULTRIX V1.2 SSB)

is:

1992-08-03-09:00:00.000 in other words yyyy-mm-dd-hh:mm:ss

with this format the record process start ok.

Some comment about this would be appreciate...

thanks in advance and regards

Julio Lerios

Are thre some news about questions in .0 y .3

I would appreciate any comments

maybe were stupid things but I would like

to know it...

thanks in advance and regards

	Sorry for the delay with this answer (I was on vacation).

	The "Historical Data Services Use" manual says that time arguments
	must be entered as an absolute time. 
	The time format for VMS and Ultrix systems are different. This is
	described in "DECmcc Use" manual page 5-18. Also in Installation Guide
	you can find how to choose default time mode.

	An example in the "Historical Data Services Use" manual is done for VMS.

	SAm