LDF file starts growing really fast

[nirvald]

We have a problem where the LDF-file for NPM database suddenly started growing in size really quickly (see graph below)

I think it started growing when we added sflow to our installation...does this make sense?

I have found the instructions on how to shrink the database, but am a bit concerned since the recovery mode was already set to simple.

/ Morgan

Orion NPM 9.1 SP4

[mcbridea]

It could be the sflow. I suggest you open a ticket on this one

[nirvald]

Done!

Case #102689 - "LDF file growing rapidly"

- Morgan

[nirvald]

And here's another funny thing, the shrink actually made the database larger!

	6/19/2009 1:30 AM		NetFlow Database Maintenance : Deleted 0 expired endpoints in 0,00 seconds.
	6/18/2009 1:30 AM		Scheduled shrink performed. DB size before shrink 5934,5625MB, DB size after shrink 6032,8125MB, released space -98,25MB.
	6/18/2009 1:30 AM		NetFlow Database Maintenance : Deleted 0 expired endpoints in 0,00 seconds.

[swack]

What was the vendor's suggestion for this problem? I'm seeing some similar behavior.

[bshopp]

What version are you on?  Do you have the DB on the same box as Orion or separate server?

In order to reduce your LDF file size, you need to stop all of your Orion services, and run a full backup of the database. Then run a Shrink on the database. 

[Chandru]

Hi Bssopp, I have the same issue with Orion netlfow where the LDF file keeps growing I did raise a support case with Solarwinds and Solarwinds support guys said issue with SQL server

SQL DBA claim that this is issue with netlfow database or tables

Can you please help us with this issue as i am not able to run any netflow reports?

Is there a way that i could run report for a specific device or IP address to check what is the high utilisation on the circuit

I have configured alerts which would trigger if the high TX or RX is more than 85% and reset when it is below 70%

i could see lots of alerts and i would like to know what is the best way of troubleshooting with netlfow enabled on the router and riverbed devices

I would like help from some one in Thwack or Solarwinds who could explain and show me some demo in detail?

Please help.....

regards

Chandru

[swack]

Netflow Developer sent me a hotfix package today to address this.  I installed it a few hours ago and so far the LDF file is only up to 1.5GB. Before the hotfix, it would grow to 65GB in a matter of hours, so I think this is working for us.

If you continue to have this problem, contact support and refer to my case (119612). He told me the hotfix was specifically for NTA 3.5 SP2.

Thanks,

Swack

[Chandru]

Hi,

I have installed NTA 3.1 SP1 and could still see the same issue

Can you guys confirm if you have a patch for 3.5SP1?

regards

Chandru

[bshopp]

Do you have a support case open for this issue?  If not, please open one.

[swack]

I have a case open with Solarwinds support (119612) and so far they've had me upgrade to NPM sp4 and NTA sp2.  After doing that, I shutdown Netflow service and waited a while, and the .ldf file was only using .5% of it's size (it had grown to 46GB).  After shrinking to 5GB and running without Netflow for a few hours, the size had not increased significantly.  However, I turned Netflow back on and within hours it was back up to 30-40GB (99% used).

My DBA pointed this out to me:

"SELECT * FROM NetFlowSummary3
WHERE [Interval] >= @StartInterval AND [Interval] <=@EndInterval"

is taking forever and needs to be optimized, it is not using a clustered index.

The support engineer has escalated the case to development team.  Any other ideas how I can optimize the configuration of Netflow in our environment?

[bshopp]

Have you looked at this thread here