When meeting with potential new customers, I’m frequently asked “do you have customized polling”. As we all learned in life, the question is often less telling than WHY the question was asked.
I answer: “Of course, we have customized polling – what enterprise class solution (or even mid-range tool) wouldn’t?”
But the next part of the conversation, when I asked “why do you ask?” usually reveals four different backgrounds to the question:
- They have been using an open source tool which is little more than a scripting engine pointed at network management. Effectively, they are developing their own network management tool one function and one script at a time. Some companies like to do this. Certainly their network ops people do – they are tinkerers. As such, tinkering with polling intervals is fun and fills their day. And if they have a small network, then it isn’t a huge task to individually set the polling invervals for a few dozen devices. Enterprise customers, with tens of thousands of network devices to manage, do not have the time or means to do this.
- They have been using a mid-range, departmental level tool, and they have hit scaling issues. What they are really saying is: “With my current tool, I have to ratchet back the polling to get it to support up to a couple hundred devices. I assume I’ll need to do that with any tool, so can I do it with yours?” Thus ensues a conversation about scalability – a single (physical OR virtual) Entuity server can support up to 1,000 devices or 70,000 managed interfaces, and you can deploy multiple servers that interoperate to linearly scale from there. You do not need to ratchet back your polling to 20 minutes (which effectively renders the monitoring useless) in order to support your, what to Entuity, is a small to mid sized network.
- They are a company who has specific devices that are extremely important, and they wish to set tighter polling intervals so further ensure availability and performance levels. The network size is less important than it criticality. They don’t want to change the polling intervals on every device. On large networks that would be untenable; they expect the NMS to have reasonable and sensible polling defaults set, but they’d like to override them on a dozen or so.
- They are a ‘super-user’ who wishes to poll specific parameters on a device that lower end tools (or even many higher end solutions) don’t poll out of the box. This is information found in MIBs not examined by default.
Entuity’s answer to all of these is ‘yes’. But the reason for the question usually exposes why the prospect is looking to replace their current solution.