Community
- Command Central
- MVP Program
- Monthly Mission
- Blogs
- Groups
- Events
- Media Vault
Products
- Observability
- Network Management
- Application Management
- IT Security
- IT Service Management
- System Management
- Database Management
Content Exchange
- SolarWinds Platform
- Server & Application Monitor
- Database Performance Analyzer
- Server Configuration Monitor
- Network Performance Monitor
- Network Configuration Manager
- SQL Sentry
- Web Help Desk
Free Tools & Trials

SQL Server running low on space

Hey all, I'm running NPM 10.2.2 with 47 Nodes, 87 Interfaces and 66 volumes being monitored. I had my SQL guy help me get NPM pointed over to our SQL box and now we're running low on drive space on the SQL server. Deke checked and SolarWinds is using 29 gigs for the SolarWinds database and 22 gigs for the log files. We've backed up SQL thinking that would erase the log files but they're all still there. I changed the retention scheme so that I'm only keeping 180 days of history but we're still running low on drive space. Is there anything else I can do that would help conserve drive space on the SQL server?

Thanks,

Joe B

Find more posts tagged with

Accepted answers

mdriskell

TrapVarbinds is the table utilized for SNMP trapping. One method if you truly aren't using SNMP trapping is to disable the service (sorry I'm at home and don't have the actual service name handy). It sounds to me like one or more of your devices is configured to send traps to solarwinds. If this is not desired I would suggest to shut off sending of traps to your server completely.

This same thing happened to me years ago when an OSPF flood caused so many traps that my DB grew from 5GB to 45GB in less than a week. The devices were mistakenly configured to send traps to SW even though I was already polling for the same data and it wasn't needed.

All comments

mdriskell

Ask your SQL guy what table is filling up the most? That seems awfully big for only 47 nodes. For instance I have 18000 elements and I'm nowhere near that large of space used.

Are you doing SNMP trapping or Syslogging? Those can contain large amounts of text that could easily begin to fill the tables.

jbruyet

Hi Mike, I don't recall configuring any Syslogging or SNMP Trapping when I was setting up my NPM but that doesn't mean I didn't. Where would I look to see if those things have been configured?

My SQL guy just checked and the table that's filling up is TrapVarBinds. If I AM doing SNMP Trapping how do I turn it off?

Thanks,

Joe B

mdriskell

netlogix

Oh, switch the DB to simple not full recovery model. Unless you are doing SQL replication, and in that case, shorten you time window between log shipments or check to see why the target or source sql server is not letting go of those logs. In either case, right click the database and shrink it, you should get most of those 22 gb back.

you should check this post Byron did about syslog, same principle applies to snmptraps.

Or just stop the solarwinds snmptrap

pacetti

Joe B,

Your problem sounds like the issue addressed in the SolarWinds Knowledge Base article, "SQL is reporting that my Orion NPM database is currently out of space. How can I clear up some space?" (http://knowledgebase.solarwinds.com/kb/questions/1673/).

HTH,

jbruyet

Hi Andrew Pacetti, I checked SysLog and Total Space Usage is 0 bytes.

Hi netlogix, my SQL guy has already shrunk the... I don't know if it was a table or a database but we went from 22 gigs for TrapVarBinds down to 9.2 gigs. I don't remember the name of the other... table/DB that he shrank. This sounds like a temporary fix because even if it was shrank it should still grow again unless I can get the problem figured out.

And I think I did that. Thanks Mike Driskell and netlogix, I saw that my SolarWinds Trap Service was set to auto and was running. I just changed it to manual and stopped the service. I'll keep an eye on things and see if this helps with my disk space problem. I see the SolarWinds Syslog Service is running but since my SysLog database showed 0 bytes I just left it alone.

Thanks for the help everyone,

Joe B

netlogix

no problem, but make sure your SQL guy changes the DB to Simple unless he *knows* without a doubt that he *needs* full. Also, somewhere in SolarWinds documentation, it says they don't support the use of Full and recommends simple. see KB on it.

jbruyet

Thanks netlogix, I just sent him an email asking him to look into it.

Thanks,

Joe B

mdriskell

Joe,

Keep in mind that even if the service is disabled the messages are still transmitting through your network to your server. If you don't utilize trapping period I would disable the node(s) that are sending it in. Like an access list you want to stop the traffic as close to the source. If you don't a future upgrade of SolarWinds or user error could cause that service to be re-enabled putting you right back into this mess.