vCenter Server monitoring with Netdata

VMware vCenter Server is advanced server management software that provides a centralized platform for controlling your VMware vSphere environments.

This module collects hosts and vms performance statistics from one or more vCenter servers depending on configuration.

Charts

It produces the following charts:

Virtual Machine

• Cpu Usage Total in %
• Memory Usage Percentage in %
• Memory Usage in KiB
• VMKernel Memory Swap Rate in KiB/s
• VMKernel Memory Swap in KiB
• Network Bandwidth Total in KiB/s
• Network Packets Total in packets
• Network Drops Total in packets
• Disk Usage Total in KiB/s
• Disk Max Latency in ms
• Overall Alarm Status in status
• System Uptime in seconds

Host

• Cpu Usage Total in %
• Memory Usage Percentage in %
• Memory Usage in KiB
• VMKernel Memory Swap Rate in KiB/s
• VMKernel Memory Swap in KiB
• Network Bandwidth Total in KiB/s
• Network Packets Total in packets
• Network Drops Total in packets
• Network Errors Total in errors
• Disk Usage Total in KiB/s
• Disk Max Latency in ms
• Overall Alarm Status in status
• System Uptime in seconds

Configuration

Edit the go.d/vsphere.conf configuration file using edit-config from the Netdata config directory, which is typically at /etc/netdata.

cd /etc/netdata # Replace this path with your Netdata config directory
sudo ./edit-config go.d/vsphere.conf

Needs only url, username and password. Here is an example for 2 servers:

jobs:
  - name: vcenter1
    url: https://203.0.113.0
    username: admin@vsphere.local
    password: somepassword
    host_include: [ '/*' ]
    vm_include: [ '/*' ]

  - name: vcenter2
    url: https://203.0.113.10
    username: admin@vsphere.local
    password: somepassword
    host_include: [ '/*' ]
    vm_include: [ '/*' ]

For all available options please see module configuration file.

Hosts/vms filtering

Module supports filtering hosts and vms. Filtering options are host_include and vm_include.

• host_include is a list of match patterns: /Dc pattern[/Cluster pattern/Host pattern].
• vm_include is a list of match patterns: /Dc pattern[/Cluster pattern/Host pattern/VM name].

Pattern should start with /. It matches name, syntax: simple patterns.

Examples:

    host_include: # filter all hosts
      - '/!*'
    vm_include: # allow all vms
      - '/*'

host_include: # allow all DC1 datacenter hosts and DC2 datacenter hosts except HOST2
  - '/DC1/*'
  - '/DC2/*/!HOST2 *'
vm_include: # allow all vms from datacenters whose names starts with DC1 and from all hosts except HOST1 and HOST2
  - '/DC1*/*/!HOST1 !HOST2 */*'

Update every

Default update_every is 20 seconds and it doesnt make sense to decrease the value. VMware real-time statistics are generated at the 20-seconds specificity.

It is likely that 20 seconds is not enough for big installations and the value should be tuned.

To get better view we recommend to run the collector in debug mode and see how much time it will take to collect metrics.

Example (all not related debug lines were removed):

[ilyam@pc]$ ./go.d.plugin -d -m vsphere
[ DEBUG ] vsphere[vsphere] discover.go:94 discovering : starting resource discovering process
[ DEBUG ] vsphere[vsphere] discover.go:102 discovering : found 3 dcs, process took 49.329656ms
[ DEBUG ] vsphere[vsphere] discover.go:109 discovering : found 12 folders, process took 49.538688ms
[ DEBUG ] vsphere[vsphere] discover.go:116 discovering : found 3 clusters, process took 47.722692ms
[ DEBUG ] vsphere[vsphere] discover.go:123 discovering : found 2 hosts, process took 52.966995ms
[ DEBUG ] vsphere[vsphere] discover.go:130 discovering : found 2 vms, process took 49.832979ms
[ INFO  ] vsphere[vsphere] discover.go:140 discovering : found 3 dcs, 12 folders, 3 clusters (2 dummy), 2 hosts, 3 vms, process took 249.655993ms
[ DEBUG ] vsphere[vsphere] build.go:12 discovering : building : starting building resources process
[ INFO  ] vsphere[vsphere] build.go:23 discovering : building : built 3/3 dcs, 12/12 folders, 3/3 clusters, 2/2 hosts, 3/3 vms, process took 63.3µs
[ DEBUG ] vsphere[vsphere] hierarchy.go:10 discovering : hierarchy : start setting resources hierarchy process
[ INFO  ] vsphere[vsphere] hierarchy.go:18 discovering : hierarchy : set 3/3 clusters, 2/2 hosts, 3/3 vms, process took 6.522µs
[ DEBUG ] vsphere[vsphere] filter.go:24 discovering : filtering : starting filtering resources process
[ DEBUG ] vsphere[vsphere] filter.go:45 discovering : filtering : removed 0 unmatched hosts
[ DEBUG ] vsphere[vsphere] filter.go:56 discovering : filtering : removed 0 unmatched vms
[ INFO  ] vsphere[vsphere] filter.go:29 discovering : filtering : filtered 0/2 hosts, 0/3 vms, process took 42.973µs
[ DEBUG ] vsphere[vsphere] metric_lists.go:14 discovering : metric lists : starting resources metric lists collection process
[ INFO  ] vsphere[vsphere] metric_lists.go:30 discovering : metric lists : collected metric lists for 2/2 hosts, 3/3 vms, process took 275.60764ms
[ INFO  ] vsphere[vsphere] discover.go:74 discovering : discovered 2/2 hosts, 3/3 vms, the whole process took 525.614041ms
[ INFO  ] vsphere[vsphere] discover.go:11 starting discovery process, will do discovery every 5m0s
[ DEBUG ] vsphere[vsphere] collect.go:11 starting collection process
[ DEBUG ] vsphere[vsphere] scrape.go:48 scraping : scraped metrics for 2/2 hosts, process took 96.257374ms
[ DEBUG ] vsphere[vsphere] scrape.go:60 scraping : scraped metrics for 3/3 vms, process took 57.879697ms
[ DEBUG ] vsphere[vsphere] collect.go:23 metrics collected, process took 154.77997ms

There you can see that discovering took 525.614041ms, collecting metrics took 154.77997ms. Discovering is a separate thread, it doesnt affect collecting.

update_every and timeout parameters should be adjusted based on these numbers.

Troubleshooting

To troubleshoot issues with the vsphere collector, run the go.d.plugin with the debug option enabled. The output should give you clues as to why the collector isn't working.

First, navigate to your plugins directory, usually at /usr/libexec/netdata/plugins.d/. If that's not the case on your system, open netdata.conf and look for the setting plugins directory. Once you're in the plugin's directory, switch to the netdata user.

cd /usr/libexec/netdata/plugins.d/
sudo -u netdata -s

You can now run the go.d.plugin to debug the collector:

./go.d.plugin -d -m vsphere