Questions about 3.0

[rbradberry]

I used version 2 for about a month but it ate up all of my system's resources (Postgre SQL took up almost 1.5G of ram to itself)

I see that the new version was released and I have a few questions:

Can I use a MSSQL Backend? If so then how can I do it.
I know it can monitor some Cisco Appliances but can it monitor my Fortigates?
I use Merak Mail Server, how can I monitor that?
I want redundancy so how can I install this on 2 servers and not have them conflict with eachother (i.e I don't want 2 messages if a server goes down)
Was the problem with the lengthy SMS messages resolved? (This was one of my deciding factors of taking it offline)
I notices the monitoring is almost all process based (i.e. if a process does not respond it alerts) can I set up periodic Pings and other network commands to test things like VPN's?

Thanks for all the help.

[cwitt_hyperic]

> Can I use a MSSQL Backend? If so then how can I do
> it.

Currently, no. PostgreSQL and Oracle are the only two supported databases for the HQ backend.

> I know it can monitor some Cisco Appliances but can
> it monitor my Fortigates?

If they are open to SNMP query, you can get availability and uptime out using the generic SNMP service via a platform running an HQ agent. More specific information for these devices may be possible through use of a custom plugin. Check-out the Plugin Develoment Center (http://support.hyperic.com/confluence/display/DOCSHQ30/Plugin+Development+Center) for some ideas on how one might do this.

> I use Merak Mail Server, how can I monitor that?

If you just need to monitor IMAP, POP3, or SMTP basic statistics and availability, you can make use of the available network service checks via a platform with an installed agent. More specific data could be gathered by way of a custom plugin. My favorite example of how this might work is our qmail plugin at http://support.hyperic.com/confluence/display/DOCSHQ30/qmail+Plugin.

> I want redundancy so how can I install this on 2
> servers and not have them conflict with eachother
> (i.e I don't want 2 messages if a server goes down)

Currently, there is no fail-over option to create redundant HQ Servers. What others have done to answer this need is to have a dedicated HQ installation (on lesser hardware) to monitor the production HQ Server, and its data connections. This way you can at least know of an HQ failure in a timely fashion should one occur. Since the agents collect and spool their data locally, they will send it to the production HQ Server once it becomes available again, so data should not be lost.

> Was the problem with the lengthy SMS messages
> resolved? (This was one of my deciding factors of
> taking it offline)

This has been answered in HHQ-183 (http://jira.hyperic.com/browse/HHQ-183). If you use standard notification, the SMS notifications contain identical content to email notifications. If you make use of the new escalation features in 3.0.x, SMS messages are less verbose, and better match the intended use.

> I notices the monitoring is almost all process based
> (i.e. if a process does not respond it alerts) can I
> set up periodic Pings and other network commands to
> test things like VPN's?

If you look in the [agent_install_path]/pdk/examples directory, you will find an example ping plugin. Follow the Plugin Deployment instructions on the Plugin Development Center page, and then configure the ping network service on a platform with an HQ agent installed. Through examination of the Plugin Development Center examples (including ping-plugin.xml), you should be able to come up with ideas for other network services. The built-in network services you can use "out-of-the-box" are listed at http://support.hyperic.com/confluence/display/DOCSHQ30/Network+Services.