To: | nv-l@lists.tivoli.com |
---|---|
Subject: | Re: [nv-l] Object Status |
From: | jshanks@us.ibm.com |
Date: | Thu, 11 Jul 2002 07:16:51 -0400 |
I am talking about a smartset of all your interface level objects. If they are named something like <router_name>:fan then your opertaors could see which routers have internal problems very easily. The idea is just like what nvsniffer creates. James Shanks Level 3 Support for Tivoli NetView for UNIX and NT Tivoli Software / IBM Software Group Francois Le Hir/Quebec/IBM@IBMCA
James, Thank you for the explanation. My understanding was that the "User2" status was used to define a specific status for an object, a status different than Down and that I could use it the way I want. I think I will go with the idea of creating an object at the interface level of the routers to show the "internal" errors. And it will propagate and change the color of the routers to yellow. Our operators already have too many popups on their screen, on top of the alerts in the event window. What do you mean by "You could collect all these in smartset and monitor that as an alternative to what you are trying to do." Are you talking about a smartset that would contain every routers ? There is already such a smartset but How would it help me ? (Most of the time when we get temperature alerts, it's the air conditioning of a remote site that failed so all devices on site are affected and the geographical representation of the devices is important for the operator to "correlate" the information) Also, if I want to create such an object in all my switches/routers, would you recommend one for the temperature and one for the fan (maybe even one for each fan) or only one for each device and a specific field that I will set to contain the error that was received ? Also is there a way to create an "interface level" object automatically (ie in a script) for all the switches and routers ? (more than one thousand) or do I have to do it manually one by one ? Also I didn't receive the document you attached in a readable format as I am subscribed to the digest version of the list. Could you please sent it to me directly ? Thanks Regards, Francois Le Hir Network Projects & Consulting Services IBM Global Services Phone: (514) 205 6695 Date: Wed, 10 Jul 2002 16:31:22 -0400 To: nv-l@lists.tivoli.com From: jshanks@us.ibm.com Subject: Re: [nv-l] Object Status Message-ID: <OF0394E58E.75DE164F-ON85256BF2.006C55ED@us.ibm.com> --=_mixed 0070E1CB85256BF2_= Content-Type: text/plain; charset="us-ascii" NetView is not nearly so simple internally as you suppose. What you want to do is a laudable goal, perhaps, but the product is deliberately designed so that users cannot screw it up, no matter how well-meaning they are. It is not the trap which sets the status on the router. It is netmon which does that, in the databases, and then issues the trap to tell everyone else about it. It also sets the status of the owning node of an interface independently, and may send a trap about that, too. ipmap on the read-write map, gets the traps, and then checks the topology database, before updating the map. That is to prevent some cases of deliberate falsing, such as you are attempting. Put another way, netmon and ipmap work together to make sure that the status of the objects they monitor cannot be messed up by users playing around with traps. You can create a trap to change the map status (color) of an object at the interface level using a trap, but not at a higher level, unless you make other changes as well, and in the case of routers that is definitely not advised. Why? Because you will most likely interfere with normal processing. It would defintely interfere with RFI. And even if not, you would need another mechanism to change the mechanism back to green when the condition you are monitoring changes Remember that objects have differnet sources for their status. Interfaces have Status Source Object, but Nodes with one interfcae have Status Source Symbol, and with more than one interface (such as Routers) it will be Status Source Compound Propagated, which means that their status will be the calculated sum of their interfaces. In other words, routers don't go red unless all their interfaces go red, for anything less than that they are yellow, or marginal, unless you change that default propagation rule to "Propgate most Critical". Off the top of my head, I don't remember whether User2 propagates up or not -- that's not my area of the code, but even if it does, the most that will happen is that the router would be yellow. So there is no way for you to alter the color of the router icon itself using a simple trap. What you have done with your custom trap is similar to using the NetView status trap (58916871) to change the color (but not the topology or database status) of a map icon, in this case, most likely the interface which corresponded to the Origin of the trap you sent. Ideally, rather than changing the status of an interface you would use your trap or the status trap to chnage the status of another interface-level icon, say a Software object or a Fan object, which you created and put on the map. You could collect all these in smartset and monitor that as an alternative to what you are trying to do. Unless I am mistaken, I am the author of the document that Leslie referred to. I originally wrote it in 1997 for NetView Version 4.1.2, but what is says is still correct. I have attached it here below (this is probably its tenth time being put to the list as it gets asked for about twice a year I think). But I would shy away from trying to turn the routers different colors. If you want to alert operators, I'd use a popup message or e-mail sent by ovactiond or in a ruleset. If you want something on the map, then creat your objects. But levae netmon's alone or you are just asking for trouble. James Shanks Level 3 Support for Tivoli NetView for UNIX and NT Tivoli Software / IBM Software Group Francois Le Hir/Quebec/IBM@IBMCA 07/10/2002 03:33 PM To: nv-l@lists.tivoli.com, Leslie Clark/Southfield/IBM@IBMUS cc: Subject: Re: [nv-l] Object Status Hello List, I am wondering, isn't every trap supposed to be able to change the status of an object ? according to trapd.conf man page: status A single character indicating the status that is to be assigned to an object when it generates the defined trap: 0 Defaults 1 Unknown 2 Up 3 Marginal 4 Down 5 Unmanaged 6 Acknowledge 7 User1 8 User2 So if I create a new trap in trapd.conf with a "status" value of 8, it should change the status of an object to "User2" right ? I tested it and only one of the interface of the object change to User2. How are we supposed to use that ? One of my goal would be to use "Temperature alerts" or "Fan Down alerts" to change routers to "User2" so that operators will notice there is something wrong even if they don't notice the alert among the "thousands" that scroll by on the even window. Leslie, I would like to see this "authoritative document on this subject", can you please send it to me ? Regards, Francois Le Hir Network Projects & Consulting Services IBM Global Services Phone: (514) 205 6695 Date: Mon, 8 Jul 2002 18:53:52 -0400 To: nv-l@lists.tivoli.com From: "Leslie Clark" <lclark@us.ibm.com> Subject: Re: [nv-l] Object Status Message-ID: <OFD894CBB6.EFBC8523-ON85256BF0.007CF29E@raleigh.ibm.com> Use the Status Trap: /usr/OV/bin/snmptrap \ $NVSERVER .1.3.6.1.4.1.2.6.3.1 \ $HOSTNAME 6 58916871 1 \ .1.2.6.1.4.1.2.6.3.1.1.2.0 Integer 14 \ .1.2.6.1.4.1.2.6.3.1.1.2.0 OctetString $SELECTION_NAME \ .1.2.6.1.4.1.2.6.3.1.1.2.0 OctetString "Object status is" \ .1.2.6.1.4.1.2.6.3.1.1.2.0 OctetString $STATUS where $HOSTNAME is the sender, $SELECTION_NAME is the object to change and $STATUS is Up, Down, etc. I'll send you James' log, authoritative document on this subject, since= you are going to need it eventually and it has been posted many times in the past. Cordially, Leslie A. Clark IBM Global Services - Systems Mgmt & Networking Detroit --------------------------------------------------------------------- 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) --=_mixed 0070E1CB85256BF2_= Content-Type: application/octet-stream; name="override.info" Content-Disposition: attachment; filename="override.info" Content-Transfer-Encoding: base64 T3ZlcnJpZGluZyBzdGF0dXMgdXNpbmcgYSB0cmFwICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICANCiAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAg ICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgICAgDQpPbmUgb2Yg ... --------------------------------------------------------------------- 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> |
---|---|---|
|
Previous by Date: | Re: [nv-l] server web enabled icon, Dermott |
---|---|
Next by Date: | [nv-l] Tivoli NetView - detecting flapping/flakey links, Donald Mahler |
Previous by Thread: | Re: [nv-l] Object Status, Francois Le Hir/Quebec/IBM |
Next by Thread: | Re: [nv-l] Object Status, Francois Le Hir/Quebec/IBM |
Indexes: | [Date] [Thread] [Top] [All Lists] |
Archive operated by Skills 1st Ltd
See also: The NetView Web