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

SNMP Hardware traps alerting

Although I know that the obvious answer will be unfavorable, I am still going to ask hoping there would be some smart thwackers who have already tested and implemented a workaround.

Issue: We have many types of network gear ranging from the typical router/switches/firewalls to Load Balancers, WAPs and management stations of which we need to get alerted on hardware degradation or failure. We all know that the out of the box hardware poller is not applicable for every type of devices and hence we need to rely on custom poller. To make things difficult we have another challenge and that is of the unavailability of OIDs for properly monitoring the main hardware components or most of the times getting the dreaded "OID not supported" message. This makes the hardware failure monitoring very difficult and as the Murphy's law goes - the 1 device you did not monitor will fail and cause an outage and you have to explain what were you doing for the past 2 years in setting up monitoring

Question is: How can we setup SNMP trap to alert on any kind of hardware failure across the number of different vendor and their network gears? Do we need to setup the trap rule individually for each type of devices or there can be a generic one which only covers any type of hardware failure or degradataion? If it has to be created per vendor or model then it will be very difficult to identify the rule criteria because you have to wait for the trap to show up once. thi smay not always be possible so what is the best way to get reported for any hardware failures in an ideal world (which is actually not ideal) !

I am sure we all have this problem and are trying to get an answer,

Thanks in advance!

Find more posts tagged with

Accepted answers

All comments

RichardLetts

Please search for my previous posts on how to generate Alerts based on SNMP traps.

Got that? okay assuming you've found one where I talk about coloring the set and reset traps I'm going to move on from there...

In this case you need to be able to generate a test trap of the right type -> some equipment vendors have a test function for this, or you can download and install the NET-SNMP package and can use that to generate traps using it for testing.

In the trap viewer you are going to normalise a number of hardware failure messages into one color for the hardware-failure type traps, and also copy the key data you need for the alert into the TAG.

Now, how you do it is going to depend on the equipment and the traps it generates -> this is the general outline of a solution for the general problem you stated. (you are going to have to experiment a bit, and open support cases.)

-> we really need to put pressure on SolarWinds to improve their support in this area because after 4 years it's getting tiresome to have to repeat this and support cases is one way to do that (another is to vote up ideas on thwack)

/RjL

alok.malvi

Thanks for the reply and direction Richard.

So essentially it does come down to getting a test hardware alert and add it manually over time.

As mentioned in my original post, its not only pain but also comes with other disadvantages like:

- maybe the test alert captures a broad category of trap type (ex: NotificationType:7). Eventually there will be many more unnecessary traps under that category. Imagine that your messages are tied to a ticketing system and type7 shoots emails for memory full or something like that.

- Maybe the test applies to a typical version and, as we all have environments which have multiple network gear and multiple version for the same model, the traps may vary resulting in some traps being missed when it really counts.

I will leave the thread open so we can pick some other brains or bring this to Solarwinds' attention.