I created a platform for the cisco IOS box --whats the best way to monitor

[hypermike]

I created a platform for the cisco ios box --whats the best way to monitor/ alert for interfaces that go down.. If i click on any one interface I get the alert tab and can create 1 alert for 1 interface .. Is there a recommended way to monitor these many interfaces..? Riunning 4.1 open source version

[admin]

Mike,

This is a case where some of the features of the Enterprise Edition would be helpful. Here's why

* You can define a resource type alert with EE. This means that you define 1 alert for the resource type associated with the Interface. It then creates alerts for each individual interface (saves time) This feature has some draw backs, however, for most use cases it works well.
* There's a hierarchical alerting mechanism in EE. This means that if the device itself goes down you won't be bombarded with alerts on each individual interface plus the device. (Like in the case of a 48 port switch. You won't get 49 alerts all at once) (This feature was added in 4.2 EE)
* There's the platform dependency manager. Which extends on the second bullet by allowing you define which agents are behind that switch. If the switch goes down you aren't hit with the 49 alerts from the switch + 48 more alerts for each agent.(This feature was added in 4.2 EE)

With that being said with the Community Edition your best hope would be to use HQApi to define the alerts in batches. This isn't a very pretty method and does require some knowledge of XML. There's some basic reference documentation for this method available at http://support.hyperic.com/display/DOC/HQApi+alertdefinition+command

Unfortunately that documentation is just basic reference and doesn't really give a good step by step guide on how to set things up.

As I primarily deal with the EE version and the resource type alerts I'm not too familiar with using HQApi to create the individual alerts. Hopefully there are others on these forums who have done this. I know we have at least one person who has and I'm hoping he reads this and can offer some tips for you 🙂

[hypermike]

Can I use admin -helper to do this multiple alerts setup thatyou say EE makes easy ..

Also We have had some datacenter issues with colling and power (put in new power substations , rewiring that stops air/coolers etc.. Is there a way not just for cisco boxes to use snmp to monitor
other servers like Solaris,linux ..for inlet/temp etc

DOES EE do trending ..so that I can save the temp,humidity, etc data .

As always thanks for your reply !!

[admin]

Mike,

I've never used the admin helper plugin. I do know that it doesn't work well with newer versions of Hyperic.

You can develop a plugin to monitor anything that is capable of providing you the data you need. If the devices provide the information you are looking for via SNMP or another means it is definitely possible to retrieve this. The caveats are that you will most likely need to develop the plugin yourself. If the means to retrieve the information doesn't use standard protocols (such as SNMP) then you'll need to develop that also.

So the answer is yes but mainly depends on what the device you are wishing to monitor does.

[hypermike]

Does EE do trending and how long can I save certain data that I would like to trend.

If someone buys the EE version do software/support write plugins for us/ or is it up to us to write it and get help from software support.

Can EE prduce weekly/daily reports ?

[admin]

Both versions store detailed historical data for no more than 7 days. EE introduces a concept of baselines http://support.hyperic.com/display/DOC/Baselines

There isn't really an ability to see trends outside of what you can visually see on the graphs. Baselines will show you the "norm" for metrics, however, they are not recalculated constantly due to the expense of resources required to calculate baselines.

There are a few options for custom plugins. As with the community edition you can seek help in these forums, you can get developer support which provides more development type questions support, you can also ask for a new feature (you can do that here in the forums with the community edition also) and there's also professional services which will basically develop the plugin for you for a fee. Normal support doesn't provide custom plugin development.

There are some reporting features in EE for graphs. They are predefined and not customizable. Here's an overview of what is possible: http://support.hyperic.com/display/DOC/Reporting+Overview

There was an add on product called IQ however it has been end of lifed.

You can download a 30 day eval of EE at http://www.springsource.com/landing/hyperic-enterprise-edition-download it's fully functioning and other than using the builtin postgres db (MySQL is recommended for performance reasons) makes a great way of seeing all the features of EE.

You also receive a support contract with EE that allows you receive prompt production systems support.

There are some other options you can explore. However I'm not the best person to discuss these if you are interested you can contact a Sales Representative and discuss all your concerns.

[hypermike]

Your John Walker gave a web cast a while back ( now in archives that I watched today). John used to work with my boss Brian Finley years ago.. In that webcast he shows integration of Nagios into hyperic. Currently the Cisco ios boxes are being monitored by Nagios . My new plugin you helped me with now monitors the ENV metrics of the cisco box and is working nicely.. Using the open source however , I cant easily filter alerts for all the cisco box interfaces in case the whole box goes down ( to just get 1 alert and not for all the interfaces on the many interfaces (using Hyperics built in cisco-ios plugin) .. Since the network group is using Nagios to filter out multiple alerts etc , I was thinking of incorporating Nagios into my hyperic server.. My question is to do this do I need to install an agent on the nagios server. It wasnt clear from what I saw if thats what I need to do .