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
Store

Any tips on how to improve speed of server?

Hello,

We are due to upgrade our Solarwinds server in three months time but in the meantime our current server is crawling along. Can anyone give me any helpful tips or direct me to documentation that will give us some performance back.

We currently have:-

Orion Platform Windows Server 2008 R2 SP1

OS Version 6.1.7601.65536

MS SQL Server 2008 R2 (SP3)

NPM ver 12

SAM ver 6.2.4

Find more posts tagged with

Accepted answers

All comments

jm_sysadmin

I wish there was an easy answer here, I struggle at times as well. I don't have a lot of info from your post about how many things you are monitoring, so my suggestions are vague as well.

First, pull out the administrator guide for both SAM and NPM, take their requirements for your size and add them together (OK, truthfully its not exactly one to one, but its a place to start).

Next, if your database is on the Orion server, consider moving it. You have SAM so using the SQL AppInsight monitor on the Orion database will give you good info if you need to give your database some love.

Lastly, look at all the standard metrics, and ensure that the data maintenance that normally runs at 2 am is running. Also, some things in version 12 were "slow" for some people, 12.1 has made me happier. Between 12 and 12.1 I had to remove the dependency components to made node and application detail pages load faster.

nks7892

I would like to add one more point here in addition to those "jm_sysadmin" mentioned i.e.-

Look for the polling rate also if you are using SAM and more to it using "WMI" connections. then probably there will be load on the server due to high TCP/IP connections. You may then go for additional polling engine also put a condition to restrict the TCP/IP connection active time.

stephen.black

Going to agent based polling for Windows servers seems to take some load off the system, especially if you are using WMI as defult polling. Agent monitoring used to be a dirty word in my world but its been one of the bigger improvements to the SW toolbox in the last while.

anitazuri

Thanks for your reply, our database is on a separate server however I've since discovered that there has not been any maintenance done in a long time, I got our DBA to take a look today and he's made some tweaks. We are also looking to change some of the disks. We are using RAID5 for our data and are looking to swap these to Raid1+0

After database tweaks and the change of disks hopefully performance will improve enough to see us through to the upgrade. I will also check out the polling rate and will find info on agent based polling; this is all pretty new to me !

Thanks again

Anita

mesverrum

Without a doubt the database is by FAR the most common bottleneck that I see in client sites. Using the wrong raid types is super common and is an absolute killer for the database. Another one that your DBA likely already saw is that many people forget to set up any kind of SQL maintenance, such as defragging the indexes. Outside of the database I find that extending the idle timeout on the NPM server from the default of 20 minutes can be helpful to stop IIS from shutting down worker processes.

1. Control Panel> Administrative Tools> IIS Manager

2. Expand out Local Computer

3. Select Application Pools

4. Right Click on SolarWinds Orion Application Pool and select Properties (or Advanced Settings for Server 2008)

5. In the Idle Time-out (minutes) box, type a number of minutes (you can set to 1440 for 1 full day), and then click OK.

-Marc Netterfield

Loop1 Systems: SolarWinds Training and Professional Services

LinkedIN: Loop1 Systems
Facebook: Loop1 Systems
Twitter: @Loop1Systems

anitazuri

Thank you for your help Marc. We are using the wrong raid type so have raised a change request to remedy this.

We will be upgrading soon so this is just to improve performance until it happens, the main issue is when using the Web interface its ridiculously slow.

tallyrich

We are currently evaluating similar things. Our previous architect wanted to monitor everything on everything - "in case" we ever needed the statistic. With 11000 switch interfaces monitored I'm arguing that we don't need to monitor everything. We have a new architect and I'm in discussions with him. I can make a case for monitoring only what you need to and only what you will actually do something about. In many environments there things that are just accepted and so to monitor them would just be a waste of resources and risk having something important missed. (If there are a dozen red lights on your dashboard that just get ignored all the time there's a good possibility that when the 13th one appears no one will notice.)

So, take a look at the what and why of your monitoring as well as the rate. For example do you really need to know every minute how much space is on a hard drive, or even every day for some devices - Yes, some may be super critical, but the point is set polling for what you really need not just the "more data is better" model.

agusst

I second this....Each Solarwinds environment I came into the person managing it felt the need to monitor everything and anything. Maybe check your monitors and see if you have some duplicate monitoring going on....such as, if you're monitoring the disk space on a VMware cluster, you don't need to monitor the disks from each node as each node in the cluster share the same disk.

hpstech

Does anyone have any thoughts on the Time-out is suggested to be '0' on many blogs, but here on thwack most folks suggest 1440?

Thanks