nv-l
[Top] [All Lists]

RE: [nv-l] Acknowledging paradigm

To: "'nv-l'" <nv-l@lists.tivoli.com>
Subject: RE: [nv-l] Acknowledging paradigm
From: "Allison, Jason (JALLISON)" <JALLISON@arinc.com>
Date: Tue, 26 Mar 2002 09:53:34 -0500
Well, you are getting the results you are getting because as I described,
you are using Acknowledge in a way that was not originally intended.
-- Agreed.  I would like to say that I have not found any details on how
Acknowledging was supposed to be used.  In the TME 10 Netview Admin Guide,
section 3-56, the only mention is that you Ack an object(s) and that you
must have a read-write map, with permissions.

One thing. A node with a single interface will not turn yellow, unless you
are setting it your self. I assume you meant red.
-- We are setting it ourselves.

You might look into changing the status propgation scheme from the
default to 'propogate most critical'. This is an all-or-nothing setting for
the entire map. If a single interface is red, the highest level to which
it propogates will also be red. This setting is under File..Modify/Describe
Map, in the ipmap settings.
-- I will be looking into this today, thank you for the recommendation.

But the effect you are trying to achieve, that is, knowing when something
else goes down after you Ack the Location Icon, is what Acknowledge
is intended to achieve, and does very well. You Ack the down interface,
and the Location Icon turns green. Something else goes down, and
the Location Icon turns yellow again. But if you Ack at the top level,
you are circumventing the normal behavior and you may get unexpected
results.
-- Yes, the problem is that we have no notification of when the status
changes from red -> yellow -> red, etc.  I do understand your point though.
As noted above, we are changing the objects status.

The release notes are installed on the system with the Help. Note that
the command to acknowledge an interface object requires object ids.
The object ids for the node and for the interface are available in the
Interface Down trap, so you could automate acknowledging, either at the
node or at the interface level for certain devices, using a ruleset and a
script.
-- Thank you for the paste!  It is exactly what I needed.

-- One last question.  I am going to be looking for a work-around for this
situation.  We are not going to be able to upgrade to 7.1 at this second, so
I need to do some damage control.  I wanted to run a quick thought by you
and the list:

Setup a OVwConfirmObjectStatusCB so that when a status change comes in,
first change the objects status to OVwNormalStatus, then change the objects
status to the new status.  This way, the object will always propigate a
cleared status, but then will propigate the new status.  So I do not get
into a recursive loop, I would have to put the stipulation in the Callback
that if the status to be set is OVwNormalStatus, let it pass automatically.
This should cause every sattus change to remove the Acked status via a
cleared propigation.

Let me know what you all think,
Thanks again for the help with my field problem,
Jason

Here's the pertinent section:
=============================================================
Acknowledged Status Import/Export Capabilities

With the existing map-based behavior for Acknowledge, operators on client
machines must refresh the map to be notified that objects have been
acknowledged or unacknowledged. The acknowledged action is map-based and
only affects the current map. Also, acknowledging objects on read-only maps
is not persistent; restarting the console clears all the acknowledged
statuses. In addition, you cannot set the Acknowledged status from the
command line.

A new global-based mode for Acknowledge is available in Tivoli NetView for
UNIX V6.0.1. When global-based mode is set, the Acknowledged status is an
object attribute. All NetView clients are notified when an object is
acknowledged or unacknowledged immediately, regardless of which map is open
and even if the action is performed on a read-only map. In addition, the
Acknowledged status can be set on an object from the command line,
permitting the transfer of the Acknowledged status to backed up NetView
databases.

By default, the behavior for Acknowledged remains map-based. However, you
can set an environment variable to activate the new global-based behavior
for individual clients. You may want the old behavior for some clients and
the new behavior for others. For instance, you may have NetView client
users who do not need to see objects being acknowledged by the network
operations center.


Global-based Mode

This section describes the global-based mode in more detail.

To activate the global-based mode, set the environment variable
NVMAPGLOBALACK equal to 1 on the NetView machine. It must be set before you
invoke the NetView console.

You can acknowledge an object using one of three methods. The first, from
the map, uses the graphical user interface. The other two methods, from an
event or through nvdbimport, allow users greater control of the
Acknowledged state outside of the graphical user interface.

>From the Map. To acknowledge a map object, select the object and choose
Monitor->Acknowledge to set the field isAcknowledged to TRUE for that
object and generate an event to update all open maps (local or NFS)
connected to the NetView server. If a NetView console running in
global-based mode starts up later, the object shows as Acknowledged.

>From an event. To acknowledge an object from the command line, type:

event -b openview -e event -a object ID
Where event is either ACK_EV (acknowledge event) or UNACK_EV (unacknowledge
event), and object ID is the object ID of the object to be acknowledged in
the OVW database. For more information on this command, see the event man
page.

This event sets the status of the specified object to Acknowledged on all
open maps where the object's status is CRITICAL or MARGINAL. It also sets
the Acknowledged field to TRUE in open NetView maps. If no NetView sessions
are running in global-based mode, this field is not set. If Netview
consoles running in global-based mode start up later, the object shows as
acknowledged.

Using nvdbimport. To directly set the Ovw database field isAcknowledged for
chosen objects, use nvdbimport. See the nvdbimport man page for more
details.
Export/Import Utility for Acknowledged Status
You can review an example of using the event-based approach to set the
acknowledged state in the sample import/export utility script
/usr/OV/bin/acknowledgeUtil.

To use the script, type:

acknowledgeUtil -export -map mapname > filename

acknowledgeUtil - import < filename

A complete description of the arguments is given at the top of the script.

Only Critical and Marginal objects are changed to Acknowledged. Attempts to
Acknowledge normal objects during the import operation are silently
ignored.
=========================================================

Cordially,

Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
(248) 552-4968 Voicemail, Fax, Pager



 

                    "Allison,

                    Jason                To:     "'nv-l'"
<nv-l@lists.tivoli.com>         
                    (JALLISON)"          cc:

                    <JALLISON@arin       Subject:     RE: [nv-l]
Acknowledging paradigm   
                    c.com>

 

                    03/26/02 08:32

                    AM

 

 




Leslie,

Thanks for the reply.  A little background on how we use Acking.

We have a "Container Icon" representing multiple nodes (where each node has
one network interface).  We use Compound Propigation to move statuses up
the
tree.  Our situation is a node will turn yellow, the yellow status will
propigate to the Container Icon yellow.  Our operators will Ack the
Container Icon (not the node that changes color).  If the node changes red,
we would like the red status to be propigated up to the container icon,
removing the Acked status.  We would like the same for red -> yellow
statuses also.  We have found that the only time the Acked status will
change is when the node changes to OVwNormalStatus.  This was confirmed by
reading that pasted section from the OVw app-default file.  If I am
misunderstanding something, please let me know.

We are in the process of getting the 7.1 media to review it for our
customer.  We are not as flexable as most with respect to change.  Our
customer is internal, thus stubburn.

Can you point me in the direction of the Release notes that discuss this
functionality?  I do know Global Acking was introduced in 6.01, but I have
not seen this reference.

Thanks again for the time you took in replying,

Jason Allison
Principal Engineer
ARINC Incorporated
Office:  (410) 266-2006
FAX:  (410) 573-3026


-----Original Message-----
From: Leslie Clark [mailto:lclark@us.ibm.com]
Sent: Tuesday, March 26, 2002 12:32 AM
To: Allison, Jason (JALLISON)
Subject: Re: [nv-l] Acknowledging paradigm


See my comments interspersed...

Cordially,

Leslie A. Clark
IBM Global Services - Systems Mgmt & Networking
Detroit





                    "Allison,

                    Jason                To:     "'nv-l'"
<nv-l@lists.tivoli.com>
                    (JALLISON)"          cc:

                    <JALLISON@arin       Subject:     [nv-l] Acknowledging
paradigm
                    c.com>



                    03/25/02 04:28

                    PM








List:

We use Ack/Unack to maintain known problems and keep top level statuses in
sync.

I just read this is /usr/OV/app-defaults/OVw
!********************************************************************
!
! If the *overrideAcknowledge resource is set to False, when
! an Acknowledged node comes up (changes to ovwNormalStatus), the
! Acknowledged status will *not* be replaced with the
! ovwNormalStatus.  The node will remain Acknowledged.
!
! If the *overrideAcknowledge resource is set to True, then if
! an Acknowledged node comes up (changes to ovwNormalStatus), then
! the Acknowledged status will be replaced with ovwNormalStatus.
!
! In other words, if you want Acknowledged nodes to have their
! Acknowledged status turned off when they come back up, set the
! resource to True.
!
!********************************************************************
*overrideAcknowledge:                  True

My questions are:

1.  Does this mean the only time the Acknowledged status will be removed
automatically (by Netview), will be when the Object status is set to
OVwNormalStatus?

>> Yes, that is the default behavior. The option to have the Ack stay
>> until you clear it was added, probably at the request of a customer,
>> a long time ago, but it is not used much in my experience. IMPORTANT
>> NOTE: The intent here was that you would acknowledge at the interface
>> level. That way propogation works all the way up to the top, making
>> the map green. Interfaces are either up or down,  so by the default
>> setting, once they come up again, the Ack is cleared. If you are
>> Acking at the node level, and some of the interfaces come up but
>> not all, the Ack on the node (I think) will probably NOT clear.

2.  Is there a way to configure Netview (OVw) so that -any time- there is a
status change for an object, the Acknowledged tag is removed?

>> That is the default behavior if you do your Acking at the inteface
>> level. This is a training issue with operators, usually.

3.  If I know the Object, can I manually (programmatically) set the objects
status to 'Unacknowledged'?

>> Yes you can. This. This ability was introduced in 6.0.1 and is
>> documented in the release notes. I see it is also covered in the
>> 7.1 release notes.

>> Another comment: Some people hesitate to Ack at the interface level
>> because the green node looks the same as a regular normal node.
>> However, the IP Status of the node is still 'Marginal' or 'Critical',
>> and the interface is still 'Critical', so they can still be shown
>> in Smartsets. I use a smartset of ProblemNodes and also one of
>> ProblemInterfaces. It is easy to see there which ones are Ack'd.

Thanks in advance,

Jason Allison
Principal Engineer
ARINC Incorporated
Office:  (410) 266-2006
FAX:  (410) 573-3026


---------------------------------------------------------------------
To unsubscribe, e-mail: nv-l-unsubscribe@lists.tivoli.com
For additional commands, e-mail: nv-l-help@lists.tivoli.com

*NOTE*
This is not an Offical Tivoli Support forum. If you need immediate
assistance from Tivoli please call the IBM Tivoli Software Group
help line at 1-800-TIVOLI8(848-6548)




---------------------------------------------------------------------
To unsubscribe, e-mail: nv-l-unsubscribe@lists.tivoli.com
For additional commands, e-mail: nv-l-help@lists.tivoli.com

*NOTE*
This is not an Offical Tivoli Support forum. If you need immediate
assistance from Tivoli please call the IBM Tivoli Software Group
help line at 1-800-TIVOLI8(848-6548)





---------------------------------------------------------------------
To unsubscribe, e-mail: nv-l-unsubscribe@lists.tivoli.com
For additional commands, e-mail: nv-l-help@lists.tivoli.com

*NOTE*
This is not an Offical Tivoli Support forum. If you need immediate
assistance from Tivoli please call the IBM Tivoli Software Group
help line at 1-800-TIVOLI8(848-6548)

<Prev in Thread] Current Thread [Next in Thread>

Archive operated by Skills 1st Ltd

See also: The NetView Web