This discussion has been locked. The information referenced herein may be inaccurate due to age, software updates, or external references.

You can no longer post new replies to this discussion. If you have a similar question you can start a new discussion in this forum.

NPM Version 8.0.18 database bloat still an issue

kesteele over 19 years ago

I am still experiencing database bloat issues with the newest version. (8.0.18 at the time of this writing)

Admittedly, I monitor a large number of objects, (1431 network nodes and 332 interfaces) which was never a problem with version 5. The database bloat problem started with version 7 and still persists. My database grows about 100,000KB per day.

Version 8 includes a new feature to allow compacting the database at night during regularly scheduled maintenance. The compaction is not working for me.

I have my "nightly maintenance" scheduled to run at 8:00AM so I can be here to monitor it. I have the new "compact before maintenance" configured to run everyday. At 7:27AM the database was 1,170,468KB. At 8:00AM database maintenance ran (I verified that it ran by watching task manager and checking the log.) At about 8:20AM database maintenance finished so I checked the size and it was 1,178,696KB.

I then shutdown NPM and ran database maintenance manually. After compaction the size was 163,888KB.

I will contact Solarwinds support about this, but wanted to share my experiences to see if others were having the same issues.

0 mjhamilton over 19 years ago

Sounds like you should be really using Orion SLX for the number of devices you are monitoring...
However.. I have both Orion and EE8... not used the Monitoring tool in 8 but guess it is SQL based??

Before the automatic maintenance tools will work you must enable the SQL Server Agent else it will not kick off on its own....

Just a thought...

Martin J. Hamilton... lost in a sea of snmp...
Cancel
Vote Up 0 Vote Down

Cancel
0 jimin8r over 19 years ago

I, too, am still experiencing the rapid growth of my database. Hopefully, they'll do something about it soon.

I also noticed that the new totals graphs, for % utilization are calculated incorrectly, showing half of the utilization.
Cancel
Vote Up 0 Vote Down

Cancel
0 kesteele over 19 years ago

mjhamilton - to my knowledge the EE8 version is Access based.

jimin8r - I have not noticed the totals graphs being wrong. Where exactly is this?
Cancel
Vote Up 0 Vote Down

Cancel
0 mjhamilton over 19 years ago

that being the case I think you should consider Orion on a SE SQL server.. this will prevent your issue...

It sounds like the EE version is for smaller networks with a few devices and low usage if it is access based...

Just my two pence worth...

Martin J. Hamilton... lost in a sea of snmp...
Cancel
Vote Up 0 Vote Down

Cancel
0 jimin8r over 19 years ago

kesteele, look under an interface, under Interface Traffic Charts and the newly added graphs at the bottom. One is Min/Max/Average Tx+Rx bps, the other is Min/Max/Average Tx+Rx Percent Utilization. Open the utilization chart, then compare it with the Min/Max/Average Percent Utilization chart that already existed.
Cancel
Vote Up 0 Vote Down

Cancel
0 kesteele over 19 years ago

Ok, I see the new graph for Min/Max/Average Tx+Rx Percent Utilization is too low.

Have you reported this to SolarWinds? What did they say?
Cancel
Vote Up 0 Vote Down

Cancel
0 jimin8r over 19 years ago

I sent them a screen print showing the two graphs side by side. Ryan, the tech support guy, emailed me later and said that he had forwarded it to the development team, and would let me know when they figured it out.

BTW, sorry to hijack your bloat post....

What have they said to you about the continuing bloat issue?
Cancel
Vote Up 0 Vote Down

Cancel
0 Bryan over 19 years ago

I have contacted them about it also and was told to purchase Orion. They told me that with the number of nodes and interfaces that I was monitoring, that Toolsets wasn't designed for my situation. I think it would be nothing for Solarwinds to port NPM to SQL. The real draw to Orion is the web enabled monitoring. I use Big Brother for the quick and dirty remote view of the up and down status of my nodes, and NPM for trend and analysis. They really need to find a way to please the customer without trying to force everyone to spend an additional 8K for another product.

Lockheed Martin
Atlas V Launch Operations
Network Engineer
Cancel
Vote Up 0 Vote Down

Cancel
0 rjenk over 19 years ago

The database issue would not be one if the scheduled compact and repair would work. I have it scheduled to run twice a week on my system and it does not work. I still have to manually run it each week...so each Monday morning I shutdown NPM, compact and restart. I don't understand why they cannot get the scheduled compact to run.

Ray
Cancel
Vote Up 0 Vote Down

Cancel
0 kesteele over 19 years ago

What I'm most upset about is the fact that Version 5 never had this issue. I could run Version 5 for 6 months before needing to compact the database.

I'm working with SolarWinds support to find out why the new "compact before maintenance" feature is not working. This is still just a band-aid though.
Cancel
Vote Up 0 Vote Down

Cancel