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

Need something a little more real-time...

We're having an issue where we bring a device up, and have to wait a long time before it shows as up in the NPM. Usually someone ends up hitting 'poll now' before it comes up on it's own.

Our current PPS is set to 30, while the recommended is 107. For statistics gathering its also set to 30, while the recommended is 35.

Polling interval is set to 60 seconds, statistics gathering, 5 minutes.

Completition is 99.4

2900 elements of which 765 are nodes, 1831 are interfaces, 304 are volumes.

SNMP Outstanding goes no lower than 60, but with the polling interval set to 60 seconds, I would have thought that it wouldn't be much more than 60 seconds before there's an update.

We need something more real-time than we're getting. Is it that my polls per second is too low? ICMP Status Polls per second - 0?

Find more posts tagged with

Accepted answers

rschroeder

If decreasing your polling and collection intervals and increasing your PPS don't do the trick, and if no one else offers a good solution, have you considered using the Bandwidth Gauges in the Engineer's Tool Set?

I build a gauge for every interface that needs frequent and granular monitoring, then choose the Historical Graphs option, and hide the gauge. I get real-time displays, with an intuitive history for the last 30-60 minutes, depending on how quickly I poll. I usually set polling for every 5 seconds, but I believe it goes as low as every 1 second. Engineer's Toolset is an excellent resource for this.

Another excellent tool is the Network Monitor in the Engineer's Toolset. I import a CSV file with all my nodes into it, and turn it on during times of upgrades or maintenance. In real time, it shows if a node goes unreachable, and will notify me with a sound and an icon color change. We keep a workstation running this in the background full-time on the boss's desk. He knows about problems before users do, with this tool.

All comments

bshopp

When you say shows up, what do you mean? Getting status or statistics?

netlogix

if the recommendation is 107 Polls Per Second, but you have set it to 30PPS, then maybe the server doesn't have the time to do the poll and is always way behind. Maybe I don't understand the PPS setting.

My understanding of the terms would be a minimum of (# of Items to Poll) / (Polling interval in Secs)}. If you have 2900 elements and want to poll them all once a minute, you would need to allow at least 50 PPS. But I still think that is a little low.

I see you included Node, Interfaces and Volumes, but what about UDnPs (?). I think they are considered "polls" too. (Not sure of internal SW stuff).

I saw, change it to 107 or 120 and see how it goes.

username82

By shows up I mean that it goes from a down status to an up status.

On the box I have 15 UDnPs. I'll try uping my PPS and see if that helps.

rschroeder