NPM Engine Status

It is doubtful that this particular tool will get much attention as the push it to migrate all user interface activities to the website. In fact aspects of this tool have already been transfered to the website. You can view these under the Admin -> Polling Engines view of the website.

As for the accurate depiction, this tool was originally part of a development debuging process and was left in the code. If you understand all of the componenets it is actually a very informative set of statititics to gauge the health of the polling service. You must keep in mind, however, that it does not take into account may of the new services and collection methods in V10 and beyond.

Cheers,
Dan

Dan, Any chance you could shed a little light on the meaning behind the ICMP/SNMP Status Polling Index and  ICMP/SNMP Statistics Polling Index. Everything else is pretty straight forward but these remain a bit of a mystery. (to your support staff as well :p)

SLXer,

The SolarWinds Orion NPM Administrator Guide provides the following explanations in the sections, "Managing Orion NPM Polling Engines - Viewing Polling Engine Status - Status Pollers" and "Managing Orion NPM Polling Engines - Viewing Polling Engine Status - Statistics Pollers":

Status Pollers

The Status Pollers group provides information about the number and rate of SNMP and ICMP device status polls the Orion NPM polling engine is currently completing, as detailed in the following table:

Statistics Pollers

The Statistics Pollers group provides information about the number and rate of SNMP and ICMP device statistics polls the Orion NPM polling engine is currently completing, as detailed in the following table:

HTH,

Fair enough, so I am monitoring 8539 icmp objects and 8539 snmp objects. 

Thanks for the assistance.

Since NPM does both an initial ICMP poll and an SNMP poll for SNMP-monitored objects, it sounds like you have a total of 4910 monitored objects on your network, none of which are monitored as ICMP-only. 

Glad to be of assistance,

This is all interesting but it still doesn't make any sense.

Look.. What is the mystery behind the numbers?

Given the following:

Elements 4910

Network Node Elements 1116

Interface Elements 3679

Volume Elements 115

Where do you get 8539 or 4910 and why would those numbers be the same for both ICMP and SNMP. 

If you trying to snmp poll icmp only nodes what kind of sense would that make.. And we have 

around 700 icmp lwap ap's..

Im just having fun here.. To be honest who cares.. I only posted here because this stumped

a senior sw engineer whom i hold in very high regard. 

Well there is no 'mystery' behind the numbers. The two counters that you reference for SNMP and ICMP reference the jobs or activities that need to be accomplished per full polling cycle for status and statistics. There are multiple different jobs for gathering all of the data for each type of element that we monitor. For example, if we look at a Node there are two jobs for gathering the status of the node(up, down, warning) and response time (milliseconds). If we look at an interface there will be a job for collecting its status, which is SNMP based and involves polling the Oper and Admin status of the interface and determining an overall status for the interface. There is also a job to collect the standard statistics for an Interface such as packets in/out, errors, discards...etc. There is a table in the database called pollers where all of the jobs are listed via a coded name for each. The stats are based on the numbers of the different types of pollers in that table.

Again, remember this was put in place to help the original developers debug the application. it was never meant to be the definitive source of poller health for the end user. The most useful number out of that screen is the SNMP Outstanding count. If it continuously hovers above 200 or reaches above 900 you have issues that need to be investigated.

And I am curious to know who the senior engineer is that was stumped :-)

Cheers,
Dan

Just to be clear (in defense of un-named engineer) There was never any confusion over any of the data that we all commonly use to validate the health of the polling engines.

My question meant to highlight the totals for those numbers that don't make any immediate sense to anyone and whose true meaning may be lost to the solarwinds engineers of old.

Never the less.. Dan I will give up this quest and accept your answer as well some sort of answer. :p

Thanks for the fun o7