drive space

Ok .. I know that this depends on the data collection and retention and all that but wanted to start somewhere and figured others using more features than I could give a little insight. I am running Orion NPM with Netflow on a dedicated server with SQLExpress running on a seperate "desktop". I am not doing any sysloging or anything "extra" at the moment and all retention is still set to the defaults. I am currently collecting data from about 100 devices.

We are about to (hopefully) move off the dedicated server and onto VM running on a faster more powerful server. At the same time we are moving off SQL express and onto a full fledge SQL server. I have been asked about how much database space I will need as this is an existing SQL instance supporting other databases and need to make sure we have enough overall space.

Today I am using a little over 2G under SQLExpress but want to start leveraging some of the other features. I have hit the 4G limit a couple times but a little house cleaning for the Netflow data corrected that. I am thinking of asking for 16G-20G bur wondered what others out there are using? More? Less? Suggestions??

Find more posts tagged with

netflow

sql

sql_express

database_space

Accepted answers

All comments

r0berth1

well, up untill yesterday, my orion install has been working fine with 15G of space for SQL but had filled it up. I then increased the HDD space to 29G and re-enabled Orion, In one day it had filled the hard drive again. I am still trying to find the upper limit of HDD space needed. I am monitoring 988 elements with NPM and have about that many in NetFlow. Dont know why 15G was fine for a year and then all of the sudden it fills 29G in one day. The more drive space the better when it comes to SQL.

bleearg13

Could be your syslog table. I've had tons of messages that I was not alerting on come in from several devices and completely fill up my DB.

r0berth1

I thought the same thing at first but could not prove it. the fix was as follows:

1. Stop all SolarWinds services.

2. Detatch NetPerfMon database.

3. Delete the 11.8gig transaction log.

4. re-attach the database.

5. Compact the database.

6. re-enable all services.

that dropped the db size from around 17.8 gigs to around 584 megs. I dont know how long that will fix the problem for but I did that at 3 A.M. this morning and at 9:10 A.M. i still have 25 gigs of a 29 gig drive available. I didn't delete any thing from the database like Netflow or anything.

My setup is as follows:

Orion 9.1 SP 5 1323 Elements, 351 Nodes, 972 Interfaces, 0 alerts, 1958 Events, 2292 Pollers, 1 Polling engine, Retain Detail 7 Days, Retain hourly stats 30 Days, Retain Daily Stats 200 Days, Retain Evens 30 Days.

VM1 = MS Windows Server 2003 R2 Standard edition, Service Pack 2, 1 Intel Xeon CPU E5345 @ 2.33GHz, 2 gigs RAM @ 2.33GHz. HDD is 19.9 GB with 15GB free. This VM runs only the Orion Website and Poller.

VM2 = MS Windows Server 2003 R2 Standard edition, Service Pack 2, 1 Intel Xeon CPU E5345 @ 2.33GHz, 2 gigs RAM @ 2.33GHz. HDD is 29.9 GB with 25.8 GB free. This VM runs onle the SQL DB for Orion.

RogerWong

Here are Microsoft's recommendations in regards to managing the SQL transaction log.

http://support.microsoft.com/kb/873235