Polling Completion much lower since upgrade to 10.2

Just shooting this out there, is anyone else noticing anything strange with the upgrade to NPM 10.2. I use to have a minimum of 99.95% for polling completion but ever since the upgrade I have been anywhere from 95.23%-99.20%. Is there a hard magic number I should be hitting around?

I had a solarwinds training session once and hte gentlemen at the time said if you ever drop below 99.70ish % then you have the chance of losing data.

** side note, I have also been having a lot of error messages pop up on my webpage since the upgrade as well **

Find more posts tagged with

10.2

Accepted answers

All comments

netlogix

have you looked to see what your polling rate % is? /Orion/Admin/Details/Engines.aspx 2nd from the bottom for me. Are you getting close? Mine completion stays at 100 and rate is 32% (I have a little environment)

david.templeton

I hope you find an answer (and share it). One of my additional pollers has bounced between 24.85 and 99.99 completion. My primary server and two additional pollers are all at 5 completion.

david.templeton

I may have found part of the problem. The .sdf files are huge. I replaced them with a copy of their blanks. It seemed to help briefly, but the problem is coming back.

From one of my additional pollers that now has a polling completion of 5:

JobsTracker.sdf: 480MB before, 406MB after
PollingController.sdf: 404MB before, 411MB after
JobEngine35.sdf: 254MB before, 267MB after

Rosieodonell

So my Polling rate is at constant 73% and the completion is still all over the place. We have about 5706 Elements with 1 poller.

david.templeton

I may be closer to a resolution now.

Updated the primary from NPM 10.2 to 10.2.1. Replaced the .sdf files on all servers. Still no luck.
Re-ran configuration wizard on additional pollers. No luck
Noticed memory usage on additional pollers was slightly higher than usual. Added more memory (virtual machines). No luck.
Tried installing all Windows updates. RAN OUT OF SPACE! The 32 GB allocations were full! Added 10GB to each. Stopped all services. Defragged - additional pollers were at 17% fragmentation, primary (larger drive, not a VM) was at 33%. Restarted all services and let it run all night.

This morning the primary is running at 99.46% completion. The additional pollers are running at 90.45%, 99.9%, and 56.87%. Not perfect, but MUCH better.

david.templeton

No luck. One additonal poller keeps dropping to 5%, the primary keeps losing its sync to the database, and reports are not updating. I've created a trouble ticket.

Tuchux

I am also having a similar issue. Upgraded to NMP 10.2 and SAM 5.0 last night. (never had a problem before) Have three polling engines.....2 physical and one virtual. the two physical boxes show a completion rate of exactly 5..never changes but the polling rates stay around 60%. The virtual polling engine is around 55% Completion rate and aroung 50% Polling rate.

vsta

Is this issue fixed in npm 10.2.2?

qbarry

Yeah - we have 10.2.1 and we're also having this issue with the MSMQ queues on an additional poller. So also interested if this has been resolved with 10.2.2

Anyone?