Having problems with your account or logging in?
A lot of changes are happening in the community right now. Some may affect you. READ MORE HERE
Acclaimed Contributor.. KAKA_2 Acclaimed Contributor..
Acclaimed Contributor..
625 views

metric from perl script using OA 12.04

Jump to solution

Dear Experts,

I have configured a simple policy which runs a command collect value and then i want to store it performance database to later represent it using graphs.

for testing i deployed this policy on my test node and a new process opcgeni was initialized on this node but it does not start.

is this not supported only on ops connector and not on agent? if supported on agent, do i have to configure additionally to have this process running?

System.txt show only "Timeout while starting component 'opcgeni'

Any help in this regard is greatly appreciated.

Thank You.

-KAKA-

0 Likes
1 Solution

Accepted Solutions
Highlighted
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

It is an agent 12 feature, but not on all OS types.  Only on the OS types that are also supported by an OpsConnector - Windows and Linux.

CP.

9 Replies
Highlighted
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

It is an agent 12 feature, but not on all OS types.  Only on the OS types that are also supported by an OpsConnector - Windows and Linux.

CP.

Acclaimed Contributor.. KAKA_2 Acclaimed Contributor..
Acclaimed Contributor..

Re: metric from perl script using OA 12.04

Jump to solution

Hi Carol,

Thanks for your response. in my case it is Solaris node. if it is Agent feature then should not it be available on all OS? is there any ER in place?

is there any other policy except Policy objects Scripts -> Method > StoreCollection or SourceCollection available for Solaris?

-KAKA-

 

0 Likes
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

the opcgeni process was added for use by the Operations Connector which is supported on Windows and Linux only.  Back at the time, OpsCx was the only UI that provided a policy editor for the policies executed by opcgeni.  Over time, these policy editors were added to OMi and supported on OA 12.x without the OpsCx, but the supported platforms for opcgeni were never extended to other OS types.

I found an existing enhancement request:
QCCR8D94464 new policy types like the contrib Example polices to generate "Event from Perl Script" do not work with Aix 7.1 on 12.03 agents.

But I dont recall anything in the Idea Exchange.  You could add it there and see if it accumulates votes from others.

As for an alternative, you could look at creating a scheduled task policy that runs embedded Perl that logs metrics to the agent's performance data store using the Perl submittal API.  An example of how to use it is here:
https://community.softwaregrp.com/t5/IT-Operations-Management-ITOM/Easily-collect-custom-metrics-using-Perl-and-visualize-in-OBM/ba-p/1675463#.XAYRpmNRdhE

I really like using this Perl submittal API and find it easier to work with than the Metrics from Perl policy type.

CP.

Acclaimed Contributor.. KAKA_2 Acclaimed Contributor..
Acclaimed Contributor..

Re: metric from perl script using OA 12.04

Jump to solution

Thanks again. i took a look at link you provided. this is more or less similar using policy object method as i mentioned but anyway thanks for the good work. :)

addition to this i have a side question. i have around 150 server where i will be storing this metric. this metric has 20 possible values and i would like to present a consolidated graph for our BU. is it somehow possible to achieve this using only OMi and Agent?

-KAKA-

PS: Compare overview dashboard is limited to chart type graph only and result is not promising.

0 Likes
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

I guess it depends what you want for "a consolidated graph".
The Performance Dashboard options are:


  • Select multiple CIs which converts the display to CI Comparison and then select the metrics to graph.  Typically used to visualise around 8 or so instances because it gets too busy in a chart to distinguish much more than that at one time.

  • Select each CI and a dashboard, then click a chart and add to Favourite.  This is not really feasible to show data for 150 servers.

  • Somehow log the data for 150 servers on 1 agent such that it is multi-instance data (instance name is the remote hostname, say).  Select the 1 node in PD and can select any instance or use parameterisation in PD to more easily select a set of instances in the chart.

 

If you want more flexibility, or you want to somehow present data for the 150 servers at one time, then you might want to create a report instead of using PD.  Eg, TopN, BottomN, or flexible layout.  Reporting could require more than just OMi and agent components.  Eg, Collect Once Store Once (COSO) to store the data in Vertica and a tool to visualise it such as Business Value Dashboard or another reporting tool that can query Vertica.  (this would need the newest OpsBridge version)

CP.

Acclaimed Contributor.. KAKA_2 Acclaimed Contributor..
Acclaimed Contributor..

Re: metric from perl script using OA 12.04

Jump to solution

Thanks again Carol. unfortunately none of these option fulfill our requirement. Option 2 may fit well if it automatically resize dashboards like Watch list but it does not.

-KAKA-

0 Likes
Outstanding Contributor.. andreask Outstanding Contributor..
Outstanding Contributor..

Re: metric from perl script using OA 12.04

Jump to solution
Hello Carol, I have a similar challenge. Customer collects ping times for 50 routers on the agent of the OMi Server itself I like to add a "is data provider" relationship between the router CIs and the Agent of Omi server. Similar to what a BSMC do. so one dashboard that is selected on a router CI should gather the data from the Agent and display the ping time. One guy of R&D in BB told me that this may work. Didnt tried this yet. Do you think this may work?
0 Likes
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

I also think it should work because, when you select a CI, PD follows the is_data_collector_for link to a node CI where it then checks if that node CI has an agent with any metrics whose instancename is the same as an attribute of the selected CI.  If that selected CI is a node CI then it is the primary dns name (not a good idea to change that mapping in PD mappings since node CI is widely used) or if the selected CI is a network device ci type then you would have more flexibility to edit the PD mappings for it to map any attribute of that CI type to ensure it matches whatever you have as the instancename in the "proxy" agent.

CP.

0 Likes
Micro Focus Expert
Micro Focus Expert

Re: metric from perl script using OA 12.04

Jump to solution

Correction: the is_data_collector_for link would be from the agent CI of the proxy node to the network device.

Other things that might help if you hit problems:

a. Set the Monitored By attribute in the network device to include OM. 

b. In PD Mappings for the network device CIT (assuming your CIs are of this type) set the CI instance to the attribute whose value matches exactly the value of the instance name in the metric data on the proxy node.

Good luck.

CP.

 

0 Likes
The opinions expressed above are the personal opinions of the authors, not of Micro Focus. By using this site, you accept the Terms of Use and Rules of Participation. Certain versions of content ("Material") accessible here may contain branding from Hewlett-Packard Company (now HP Inc.) and Hewlett Packard Enterprise Company. As of September 1, 2017, the Material is now offered by Micro Focus, a separately owned and operated company. Any reference to the HP and Hewlett Packard Enterprise/HPE marks is historical in nature, and the HP and Hewlett Packard Enterprise/HPE marks are the property of their respective owners.