|
|
| Subject: | Graphing disk space utliization, one server with multiple volumes |
| Author: | [Not Specified] |
| Posted: | 2015-01-08 12:38 |
New to Zenoss, been playing around with it for a few days now. I'm not loving the monitoring capabilities of Amazon CloudWatch and its limitations placed on the free tier of service. I wanted to kick the tires on an open platform like Zenoss Core before I decide to spend on something like DataDog that I find really easy to use.
So far I've been impressed with the product and now its time to get my dashboard built out.
Looking for help on how to build a single graph for disk space utilization in Zenoss 4.2.5 on my Linux EC2 instances. Using the EC2 zenpack and modeling against device class /Server/Linux using SNMPv3 to collect data the CloudWatch API doesn't already provide like RAM and disk usage.. Creating a data source for dskPercent to .1.3.6.1.4.1.2021.9.1.9.1 graphs fine when I specify a single volume in the snmpd.conf. When I add any additional mount points to the .conf I've only managed to get the sum utilization for all volumes as a single line. Ideally I'd have a single graph with a separate line for each EBS volume attached showing 0-100% space used.
Regards,
Garret
| Subject: | Hey Garret, |
| Author: | Matt |
| Posted: | 2015-01-08 15:04 |
Hey Garret,
I'm not familiar with the EC2 ZenPack, but in general volumes should show up as a component.
If that's not what you want, then you could write a custom script to pars the utilization of each volume into individual RRD files + use a Command Data Source to run this script.
However I don't think that's a good idea, because you'll end up with a fixed number of volumes for each device.
And usually you don't know how many volumes a device will have + this number might even change over time.
Hope that helps and let me know if you need more details about either solution.
BR Matthias
| Subject: | Appreciate the quick reply |
| Author: | [Not Specified] |
| Posted: | 2015-01-08 16:08 |
Appreciate the quick reply
The volumes do show up in Components/File Systems individually and graph fine. For my dashboard view I was just really hoping there may have been a quick way to combine them into one.
I realized the error in my original approach. When graphing both custom data points 1.3.6.1.4.1.2021.9.1.9.1 and .1.3.6.1.4.1.2021.9.1.9.2 I get the desired output. Problem with this method, like you mentioned with your possible solution, is that it isn't flexible. I have some instances with up to 15 volumes attached, so if I continued on with how I have it working now would leave the graph with a lot of legends and waste space in the graph. If I shortened the custom data point back one value the snmpwalk will output unique percentage values for as many mount points referenced in snmpd.conf but that data point doesn't graph as desired.
| Subject: | Maybe a Graph Report would |
| Author: | Matt |
| Posted: | 2015-01-08 16:54 |
Maybe a Graph Report would work!
I believe that you can create aggregated Graph Reports that include all the volumes.
You could then use the Show Graph Portlet ZenPack to display that Report on your Dashboard.
Would that work for you
| Subject: | That solution would be fine |
| Author: | [Not Specified] |
| Posted: | 2015-01-09 09:53 |
That solution would be fine by me. I'll give it a shot, thank you for the suggestion.
| < |
Previous Missing graphs with WMI monitoring |
Next Monitor a Specific VM via vSphere Zenpack? |
> |