cancel
Showing results for 
Search instead for 
Did you mean: 
Create Post

Disk Growth Monitoring

Disk Growth Monitoring

We find that we have found that some problems crop up every now and again such as Exchange issues that cause large amounts of Transaction logs to occur. We have our clients Exchange logs on separate drives. Every now and again we have an iPhone go on the fritz, an Outlook profile issue, or a PST that gets caught causing an outrageous amount of logs.

We have low thresholds set across the board for our clients mainly by request. We are alerted around 10GB of free space currently. Let's say this issue crops up. Exchange server has a 300GB hard drive in it. The drive has 100GB of free space (just as an example) on it. What we are looking for is a way to monitor that disk for size change in a specific amount of time.

If the disk usage skyrockets at any  time in comparison to it's normal disk usage it would send an alarm.

Let's say for our example Exchange eats 10GB of space every half hour, but normal usage is that it uses 1GB every hour, then it sends an alert.

This way we would potentially catch this alarm at 90ish GB free space and see the problem prior to it getting to 10GB of free space and having 30 minutes to clear the logs, and then start troubleshooting.

We are essentially looking for a way to alarm based upon disk usage. The ability to define a warning/critical global, and the ability to override this for specific drives that have high usage (Backup drives).

12 Comments
Product Manager
Product Manager

SAM 6.2 (currently in beta) includes trending and forecasting on volumes so you can be alerted that you're going to run out of space on a volume in "X" days.

Level 10

aLTeReGo I am running 6.2 Beta. The Trending is very nice, and trust me I have been waiting for 6.2 to become generally available.... It has some awesome features! 6.2 coupled with NPM 12 is great! Anyways I think that the forecasting will help, but still not solve our issues unfortunately.

Good suggestion though! thank you very much! 🙂

Level 11

YES!

Level 8

Trending is great, but what the OP suggests is the ability to alert on rate of change e.g. X% within N hours. Could be useful in catching unforeseen events.

Level 12

very nice

Level 7

Alerting an "abnormal" growth is what's required. Considering storage forecast data is captured then this data could be used.  For eg,  If normal growth us 2% per day and over N samples growth equals,  say,  10% then provide abnormal growth alert. Should/Could this be moved to SRM?

Level 10

That is a great idea, unless said persons is not using a SAN. They may be using local storage, and/or a Physical server.

I was only looking at this from a disk/virtual disk aspect.

Thanks! 🙂

Level 14

Yes,  would be a quick way to see backup failures where transaction logs don't get deleted correctly.  Monitoring the reverse where a significant drop doesn't happen.

Level 17

I know this thread has quieted down, but if you still have a need please check out my Volume Usage Spike Detection Alert

-CharlesH

Loop1 Systems: SolarWinds Training and Professional Services

I had a similar issue with Solarwinds communications filling up the local server's C drive after the last major upgrade to NPM 12.2.  In response, I built this view for my NPM front page:

pastedImage_0.png

It's easy to keep an eye on C drive space, but no longer strictly necessary since hotfixes have corrected the cause.  PM me if you need the info to build this.