Solarwinds spec.

My apologies if I posted in the wrong section, but I am unsure where to put this. I am seeking a little advice.

I have taken over maintaining and building upon an existing solarwinds solution.

The vm is (which is stand alone) windows 2k8, 12gb ram, 4 cores. I have noticed that (a) the original person who designed this made everything clunky and visually confusing (they included a lot of extra content, such as top 10 x, nodes up/down and a search engine that I feel you don't need to include in the home page) but I've also noticed that when you request anything (reports, managing alerts, ect) it's horribly slow to do anything (maybe two or three minutes to render a page).

Is it simply a case that the machine needs more resources/more solarwinds machines? does it need to be trimmed down as it's rendering every graph at the same time? Is it something associated with SQL and it's demand for all of the resources it can use? or is that a feature in solarwinds? I have included a small snippet of the home page to show you what I mean.

Any advice is much appreciated.

Find more posts tagged with

slow_performance

Accepted answers

All comments

adatole

You are probably right that it needs more resources, but a little more info would help:

How many devices are you monitoring
WHERE are those devices in relation to the monitoring server itself
what is the speed of the CPUs
What about the database specs. Sometimes the problem has nothing to do with the monitoring server, but rather the database hardware.

Let's start with those questions are work our way through.

You can also contact support and they will run through these things with you (it's why you pay for maintenance, right?).

shaun_9999

Thank you for getting back to me.

My apologies for the lack of detail, I have very little experience in monitoring and even less in solarwinds. I am using google and thwack at the moment to try and understand everything.

1 - Number of devices: About 400 devices

2 - Devices located: One site is located about 80 miles away, the other is 20 miles away. The monitoring server is situated in the office.

3 - CPU speeds: AMD opertons 6136 2.40

4 - DB specs: The specs are as follows:

Ram - 8gb, Win 2k8, 2.40 x 4 CPU.

Hope this helps.

neomatrix1217

What other applications do you have besides NPM? Here is our system guidelines for multiple applications Orion multi-module system guidelines Single application Plan your production deployment

shaun_9999

I am thinking, rather than have one server monitoring, I would like a machine at each site to do this. Plus, also, should you have a failover, you will always be monitoring.

shaun_9999

We/they use:

NTA

DPA

IVIM

IPAM

WPM

SAM

NPM

adatole

So I would DEFINITELY say that your database is the top potential slowdown. Not enough RAM and slower CPU (yeah, I know, I just called 2.4Ghz slow).

I also forgot to ask, but what kind of disk is the DB running on - how many spindles, RAID, etc.

Because this is an inherited system, there could be other issues - too many pollers, etc. But that's where I'd start.

designerfx

I would also add two questions:

1 *where* is the DB? Hopefully not on the same machine?

2: how many nodes?

CourtesyIT

How many sites are you looking at?

How many nodes at each site?

Is your SW Orion installed on the OS partition or separate?

How long ago was the server built?

Is the SW Orion and SQL in the same subnet?

I know lots of questions but just want to make sure we cover all the bases possible.

Thanks,

rschroeder

Firing up QoE's view of the SQL server hosting the Orion database may display the problem brilliantly. If that server is not included in QoE, now's the time to add it in so Orion can show you its performance.

designerfx

I wouldn't try to get too many things going at the same time. Maybe first get your DB off of your main solarwinds server, give it enough horsepower and get it built out properly as adatole mentions, and maybe come up with a plan for migration. Too much at once in too many directions will just make things more difficult. Just that step alone will take some time to get down. Much, much later you could look at additional pollers but you'd want some stability before you get to that.

shaun_9999

From a look at the NMP polling I can tell you the following:

How many sites are you looking at? - Three sites (atm I am only looking at one site at a time)

How many nodes at each site?

As a rough estimate:

Site 1 (in house) 280

site 2 (80 miles away) 160

Site 3 (20 miles away) 200

Is your SW Orion installed on the OS partition or separate? OS partition

How long ago was the server built? 01-17-2014 on 2k8 r2

Is the SW Orion and SQL in the same subnet? They are on the same server, so yes.

I know lots of questions but just want to make sure we cover all the bases possible.

adatole

I agree with designerfx's comments above - don't do too much at once.

From your last set of answers, my order of changes would be:

1) get that database onto a separate box!! This, more than anything, is impacting your performance and will alleviate the strain you are experiencing right now.

2) as part of that, increase your CPU and RAM on both systems. That includes ensuring that the disk that stores your datastore is NOT running RAID 5. Do either RAID 1 (a plain-old mirrored drive) or RAID 1+0.

Get that done, then let us know how it's going.

cnorborg

I'll also echo that comment. The single most important thing for Orion IMHO is a separate database that has good performance. Follow that up with sufficient memory and CPU on all boxes...

Plus, update everything to get it running the latest code, and if you're using NTA, make sure you have a seperate NOSQL server for it, in addition to the main database being hosted on a separate server.