Thursday, January 10, 2013

Introduction

Hi all,

This last week I've passed through some of my notes in order to decide from where to start. As I mentioned in my last post, the idea is not to repeat what is already written in the official documentation, so, if you want to know how to install the DataChannel or the minimum requirements, please check the official manuals (I know they are big...but you have to pass through it at least once :).

Before I continue, lets talk a little bit about history...

If you work with performance management for some time already, you probably heard about a product called Proviso...this is the former name of TNPM (or TNPM is its new name if you wish). In fact, after IBM acquired Quallaby, they decided to pick some other products from its portfolio and combine them under the same solution. They got Proviso, TIP, Cognos and Tivoli Netcool Performance Manager for Wireless, mixed them on the same cooking pot and produced two main dishes: TNPM wireline and TNPM wireless. In this blog we will discuss more about the wireline one.

Some other changes were done as well. IBM created two components for making the installation easier: one called TopologyEditor and another called Deployer. The idea was to centralize the configuration of all components in one place (the topology part) and allow local or remote installation from a central point (the deployer). Before, the configuration of each component was spread among many servers using config files. It was a good initiative from IBM, but I have to say that it brought with it more complexity to the process (and a lot of bugs as well).

Well, lets stop the history here...

As I mentioned earlier, I was checking my personal notes in order to find some interesting topics to write about and I found many of them:

  • Why does TNPM installs as root and is there a way to work around it?
  • How to manage users in TNPM?
  • Why SFTP doesn't work by default for bulk collections and how to force it?
  • Composite Subelements (CSE) introduction
  • Are you installing a Remote CME through a firewall? If you follow only the manual it won't work...
  • CME troubleshooting
  • BCOL troubleshooting
  • Creating Bulk adaptors, pvline format and other details
  • ...

I would like to try something here...instead of just choosing myself the topics sequence, I would like to ask you from where to start...The idea is to try to write about a different topic every week...I know this is a very new blog and probably few people know about it (you can change it sharing the link with your coworkers), but I want it to be as interactive as possible...so, please give me your opinion about which one you would like to see first or any suggestions you may have for new topics.

Au revoir!
Posted by at

13 comments:

  1. UnknownJanuary 11, 2013 at 10:29 AM

    Hey Danilo! What a good welcome surprise man!
    I'll be tunned here, and share some info (if I can)...
    I might have some additioin tips and tricks stored somewhere here, which I'll consult on demand.
    Good luck with the initiative ;)

    ReplyDelete
  2. Amit MalikJanuary 12, 2013 at 10:31 AM

    I would be intrested to know about
    pvline format and other details..

    ReplyDelete
  3. UnknownFebruary 21, 2013 at 9:33 PM

    1. I would like to know the proper procedure for decommisioning a device from TNPM.
    2. I would like to know any tips or tricks for creating thresholds....i.e. custom scripts created....etc..

    ReplyDelete
    Replies
    1. Danilo LunaFebruary 22, 2013 at 9:56 AM

      Hi Jeff, what exactly do you mean about "decommissioning a device"? We have some ways of doing that in TNPM. Do you want just to stop collecting data for that device? Do you want to completely delete it from the system? Can you give me some more details?

      Delete
    2. UnknownNovember 5, 2013 at 8:39 PM

      Hey Danilo....sorry so long for the reply.

      1. When I talk about decommissioning, what I'm asking is how is the following scenario normally handled in TNPM:
      Someone is going to be taking a system offline and they let us know, but we don't want the NOC to get hit with a bunch of alarms and TNPM is setup to phase the device out by default when the device fails 3 discoveries and is at least 7 days old. What is the best way to immediately stop the device from alerting? I figured putting a property on the devices like Monitored=1 then I could use the grouping rules to group the device into a folder called "Decomissioned Devices". Would this be the normal way this scario would be handled?


      2. To my other question about the thresholds. How do most people handle deploying a bunch of thresholds in TNPM. I was thinking about creating a script that would create thresholds on every metric, group, resource combination using resmgr and just disable them until they are needed. Would this be an acceptable thing to do? Would there be any issues with performance even if 75% of these thresholds remained disabled? How do most people deal with deploying a bunch of thresholds......any ideas / examples would be greatly appreciated.


      Thanks,

      Jeff

      Delete
  4. UnknownSeptember 13, 2013 at 11:40 AM

    Hi Danilo,

    Firstly, I would like to thank you for creating this blog. I am interested to know about the Library formulas in TNPM. Can you please explain about the following:

    1. How can I edit the Library formulas(eg: ifAvailability, deviceAvailability etc) and test the formulas for a device.
    2. There is no data for Interface Availability metric for one of the subelement, I found the below error in proviso.log.In Error message guide it is mentioned that this error indicates there is a problem in device but Can you please explain what exactly this error mean and where is the problem.

    2013.08.28-00.00.09 UTC SNMP.4.403-11838:6890 I [DL30112] CALC_ERROR RFC1213LIB, discarded ifAvailability result for SubElement 'gbayec70.mprn_If<20>' (Prev) dev upTime(s) 11379889.65 @(epoch) 1377647703, status up since upTime(s) 22523137.06 (Curr) dev upTime(s) 11380190.16 @(epoch) 1377648003, status up since upTime(s) 22523137.08 => (up) 31806726 + (down) 0 + (unknown) 4263160871 ~= (elapsed) 301 / 1 poll(s)

    ReplyDelete
    Replies
    1. Danilo LunaSeptember 16, 2013 at 6:05 PM

      Hi Venkatesh,

      By principle you should never edit a library formula but create a copy and then edit the copy. If you edit the original IBM version, it can be overwritten during upgrades. I personally never use the default formulas in my projects, even if I won't change anything in the formula. I always create a copy to be sure that it will never be overwritten by IBM.
      So, my suggestion is to create a separate formula collection branch and use the formulas from there. For instance, if I want to use the formula ~AP~Specific~SNMP~Cisco IP SLA~Round Trip~RTT I will create a copy with the following path ~[CustomerName]~Specific~SNMP~Cisco IP SLA~Round Trip~RTT and use my version in the requestEditor.

      Now, for the error mentioned, can you please let me know from which library is this formula from?

      Delete
    2. UnknownSeptember 17, 2013 at 11:41 AM

      Thanks Danilo,

      I was asking about Library formulas, normal formulas however i will create a copy and i can edit using formula editor. But, I want to know about ifAvailability(percent) Library formula.

      Normally if you open this formula using formula editor, it wont have any oids mentioned in it, but how can i see the content of this library formula.

      Regarding the error i have mentioned, the formula is using RFC 1213 Library....

      Thanks & Regards,
      Venkatesh Chowla

      Delete
    3. Danilo LunaSeptember 17, 2013 at 12:51 PM

      Hi Venkatesh,

      Now I understood your question. Sorry for that. I believe those formulas cannot be edited. You can create your own version for the ifAvailability calculation but you will have to start from scratch, or get some ideas from IBM manual (IBM Tivoli Netcool Performance Manager: SNMP Formula Language Reference page 101). I also found more information about the ifAvailability formula here (http://neil.tappsville.com/index.php/TNPM_KPI). I hope this can help you.

      KR

      Danilo

      Delete
  5. UnknownSeptember 18, 2013 at 7:46 AM

    Thanks Danilo for the link, but it don't have the information about interface Availability. However, the formula is given SNMP Formula Language Reference but just I was trying to test the formula in Formula Editor which is not possible.

    Can you please suggest about the error which I have mentioned.

    2013.08.28-00.00.09 UTC SNMP.4.403-11838:6890 I [DL30112] CALC_ERROR RFC1213LIB, discarded ifAvailability result for SubElement 'gbayec70.mprn_If<20>' (Prev) dev upTime(s) 11379889.65 @(epoch) 1377647703, status up since upTime(s) 22523137.06 (Curr) dev upTime(s) 11380190.16 @(epoch) 1377648003, status up since upTime(s) 22523137.08 => (up) 31806726 + (down) 0 + (unknown) 4263160871 ~= (elapsed) 301 / 1 poll(s

    ReplyDelete
    Replies
    1. AnonymousOctober 24, 2017 at 8:32 AM

      Hi Venkatesh,

      Have you resolved this issue? I have the same error message. Need your advise on this. Thanks.

      Delete
  6. Vikas MahajanDecember 29, 2017 at 12:25 PM

    Hi, AP.cisco_ipsla.Round_Trip_Latency report is full of blank values. Not sure where to start the troubleshooting ?

    ReplyDelete
    Replies
    1. Danilo LunaDecember 29, 2017 at 5:26 PM

      Please check: https://tnpmtnt.blogspot.ch/2013/09/no-data-in-tnpm-reports-some-things-to.html

      Delete

Subscribe to: Post Comments (Atom)