Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials

Dependencies using an HSRP Setup

Hello,

New to the site, but been using SolarWinds for quite a few years now.

First off would like to say that the Dependencies feature that was finally built in are great.

Here is my issue and question around this feature.

I am currently trying to set-up dependencies across 50-60 remote sites.

The sites are set up with 2 WAN provider routers. A T1 on the Primary and DSL on the Back-Up. The Routers are set up using HSRP for fail over.

The way the network currently routes is all Data/VoIP goes through the Primary Router. If the Primary fails all of the devices, but the VoIP gateway fail over to the DSL, and are still reachable through our MPLS cloud. The Gateway becomes unreachable to NPM and our network since the Primary router fails, and we don't route the VoIP Gateway subnet through the DSL due to no QoS ability across the DSL.

I have set up 2 Groups per each site. 1 Group for the 2 WAN Routers 1 Group for all the LAN/VoIP Gateway Devices. I have created Dependencies based on the WAN Routers as my Parent and my Devices Group as the child.

The issue I am seeing is that when my Primary fails I still get an alert for the VoIP gateway since it becomes unreachable and it's subnet is not routed through the DSL Router. I would figure that since the VoIP gateway is only reachable through my Primary router once it fails the dependency should kick in and take its status to Unreachable and NOT kick off another alert.

I think I am seeing this because I still have another Router in the parent device that shows Up even though the VoIP gateway isn't routed through this router.

Am I thinking correctly and do I need to split my WAN Routers into 2 groups and make 2 dependency rules for each site?? And will I need to create 2 groups for devices, so that the VoIP device isn't a child device of the Dependency group that includes the Fail-Over router.

Also, sorry for the quick Visio drawing created to a JPG.. Pressed for time since I am working on a full roll out in this infrastructure and haven't even started touching the SAM protion of it!!

Find more posts tagged with

advanced_alert

advanced_alert_setup_question

dependency_alerts_grouping_groups_suppression_suppress

Accepted answers

All comments

mrxinu

Hey Richard,

I think I am seeing this because I still have another Router in the parent device that shows Up even though the VoIP gateway isn't routed through this router.

You're absolutely right; NPM doesn't care how the packets normally get to a device. If you've told it that as long as one of two devices are up, everything's good, it's going to behave as if it's business as usual - sending you a Node Down alert when your VoIP Gateway goes missing. I think you're on the right track with having two dependency chains. One that includes your switch and server (dependent on the HSRP pair) and another where your VoIP Gateway is dependent on your T1 router. That way your switch and server can be unreachable only when BOTH are down, and your VoIP gateway will be considered unreachable as soon as the T1 router goes down.

rjnicholson

Steven,

Thanks for the quick response. I had a feeling that this was the behavior of dependencies. Do you think there would be a possibility of adding a statement to the algorithm at some point that would not only check the Parent device of the the listed Child, but also query the routing table of the 2nd Parent device for known subnets in the Child group, and then base its ability to mark a node unreachable on these results. This would be pretty sweet as it would cut down on my Group size and dependencies to manage per site location.

mrxinu

You laugh, but CJ @ SolarWinds hacked up some hierarchical dependencies when he was still doing professional services with me at Sigma using that very method. It was wild, but it worked. I can't remember the specifics but if he could codify it once in I bet the SolarWinds folks could do it again.