NPM 9.5 SLX SP3 Unlimited elements has a limit. Need another poller?

Presently we have NPM SLX 9.5 SP3 installed which is monitoring almost 10,000 elements.
It appears that this has been part of the numerous issues we have had with NPM, where it stops collecting data. Even though SLX states Unlimited, there appears to be a limit between 8000 and 10000.
We have been told to get another polling engine.
We have access to a NPM SL500 9.5 SP3 install from a company we acquired. Can this be used as a second poller to augment the NPM SLX? Will it be limited to 500 elements it can poll? or can it only be used as a separate NPM install?

If it can be used as a second poller, is there a special process to configure / setup?
Are there any 3rd party options for polling engine?

Your time and insight are appreciated, as the purchase of a 2nd poller at the 14,000.00 is a little ridiculous especially since the SLX is sold as "Unlimited".

Find more posts tagged with

npm_slx

Accepted answers

All comments

FormerMember

Hi there,

I do feel you pain, as I had similar problems where I had 4 SL2000 licenses which i have had to retire and buy SLX pollers due to the way we were moving with our installation. However, in theory, if you were to to get a mega server, you could possibly take the number of element further. if you also drastically reduce all fo the other work the primary poller is doing, alerting, syslog, custom polling, and reduce the polling and stats on all of your elements, i bet you could could take it up a little...

njoylif

something else to consider is to work with SW to upgrade the 500 count to SLX or something, you would certainly get SOME credit for the license.
I agree, though, I wish they had different licensing when it comes to SLX because we have 18000 pollers across 3 SLX boxes. I'd like to be able to spread out further for various reasons, but we won't buy more just for that...

JCollins

Could you list the specs for your poller and database server? Also from the about tab of NPM System Manager, list the number of nodes and interfaces you currently are running. I just crossed the 8000 mark on interfaces and this box is running at only 20% of CPU util. Do you think you may be hardware constrained or bandwidth constrained?

jeff.stewart

I think the documentation for the SLX recommends no more than 8,000 elements. We are running well over 10,000 on each of our servers.

mcampbell

Windows 2007 SP1 64Bit Xeon 2.53ghz 2 processors 8GB RAM

SQL 2008 on seperate server. (not sure on specs, but know its a very robust system)

Monitored network Elemetns: 1043 nodes, 8120 interfaces

I don't believe its a CPU or Bandwidth issue. CPU is pretty much nominal usage.

Thanks again for your insight.

mcbridea

Do you know what the storage for SQL is? We recommend RAID 10 and there are known issues on RAID 5 and some SANS.

njoylif

yea, its never been a system limitation, my system's CPU/Memory utilization have always been nominal. not sure where the bottleneck is...perhaps bus speed?
We have NPM and 2 addtl pollers all barely used (2k3 until we get 9.5 running well), SQL 2k5 EE running on Compellant SAN with no issues there, have 18000 pollers, NCM, NF, IPAM.

2 pollers run 7000 and one 4000 (extranet).