VMware vCenter Exporter for Prometheus.
Get VMware vCenter information:
- Basic VM and Host metrics
- Current number of active snapshots
- Datastore size and other stuff
- Snapshot Unix timestamp creation date
Requires Python >= 3.6
- Install with
$ python setup.py install
or via pip$ pip install vmware_exporter
. The docker command below is preferred. - Create
config.yml
based on the configuration section. Some variables can be passed as environment variables - Run
$ vmware_exporter -c /path/to/your/config
- Go to http://localhost:9272/metrics?vsphere_host=vcenter.company.com to see metrics
Alternatively, if you don't wish to install the package, run it using $ vmware_exporter/vmware_exporter.py
or use the following docker command:
docker run -it --rm -p 9272:9272 -e VSPHERE_USER=${VSPHERE_USERNAME} -e VSPHERE_PASSWORD=${VSPHERE_PASSWORD} -e VSPHERE_HOST=${VSPHERE_HOST} -e VSPHERE_IGNORE_SSL=True -e VSPHERE_SPECS_SIZE=2000 --name vmware_exporter pryorda/vmware_exporter
Only provide a configuration file if enviroment variables are not used. If you do plan to use a configuration file, be sure to override the container entrypoint or add -c config.yml to the command arguments.
If you want to limit the scope of the metrics gathered, you can update the subsystem under collect_only
in the config section, e.g. under default
, or by using the environment variables:
collect_only:
vms: False
vmguests: True
datastores: True
hosts: True
snapshots: True
This would only connect datastores and hosts.
You can have multiple sections for different hosts and the configuration would look like:
default:
vsphere_host: "vcenter"
vsphere_user: "user"
vsphere_password: "password"
ignore_ssl: False
specs_size: 5000
fetch_custom_attributes: True
fetch_tags: True
fetch_alarms: True
collect_only:
vms: True
vmguests: True
datastores: True
hosts: True
snapshots: True
esx:
vsphere_host: vc.example2.com
vsphere_user: 'root'
vsphere_password: 'password'
ignore_ssl: True
specs_size: 5000
fetch_custom_attributes: True
fetch_tags: True
fetch_alarms: True
collect_only:
vms: False
vmguests: True
datastores: False
hosts: True
snapshots: True
limited:
vsphere_host: slowvc.example.com
vsphere_user: '[email protected]'
vsphere_password: 'password'
ignore_ssl: True
specs_size: 5000
fetch_custom_attributes: True
fetch_tags: True
fetch_alarms: False
collect_only:
vms: False
vmguests: False
datastores: True
hosts: False
snapshots: False
Switching sections can be done by adding ?section=limited to the URL.
Variable | Precedence | Defaults | Description |
---|---|---|---|
VSPHERE_HOST |
config, env, get_param | n/a | vsphere server to connect to |
VSPHERE_USER |
config, env | n/a | User for connecting to vsphere |
VSPHERE_PASSWORD |
config, env | n/a | Password for connecting to vsphere |
VSPHERE_SPECS_SIZE |
config, env | 5000 | Size of specs list for query stats function |
VSPHERE_IGNORE_SSL |
config, env | False | Ignore the ssl cert on the connection to vsphere host |
VSPHERE_FETCH_CUSTOM_ATTRIBUTES |
config, env | False | Set to true to collect objects custom attributes as metric labels |
VSPHERE_FETCH_TAGS |
config, env | False | Set to true to collect objects tags as metric labels |
VSPHERE_FETCH_ALARMS |
config, env | False | Fetch objects triggered alarms, and in case of hosts hdw alarms as well |
VSPHERE_COLLECT_HOSTS |
config, env | True | Set to false to disable collection of host metrics |
VSPHERE_COLLECT_DATASTORES |
config, env | True | Set to false to disable collection of datastore metrics |
VSPHERE_COLLECT_VMS |
config, env | True | Set to false to disable collection of virtual machine metrics |
VSPHERE_COLLECT_VMGUESTS |
config, env | True | Set to false to disable collection of virtual machine guest metrics |
VSPHERE_COLLECT_SNAPSHOTS |
config, env | True | Set to false to disable collection of snapshot metrics |
You can create new sections as well, with very similiar variables. For example, to create a limited
section you can set:
Variable | Precedence | Defaults | Description |
---|---|---|---|
VSPHERE_LIMITED_HOST |
config, env, get_param | n/a | vsphere server to connect to |
VSPHERE_LIMITED_USER |
config, env | n/a | User for connecting to vsphere |
VSPHERE_LIMITED_PASSWORD |
config, env | n/a | Password for connecting to vsphere |
VSPHERE_LIMITED_SPECS_SIZE |
config, env | 5000 | Size of specs list for query stats function |
VSPHERE_LIMITED_IGNORE_SSL |
config, env | False | Ignore the ssl cert on the connection to vsphere host |
VSPHERE_LIMITED_FETCH_CUSTOM_ATTRIBUTES |
config, env | False | Set to true to collect objects custom attributes as metric labels |
VSPHERE_LIMITED_FETCH_TAGS |
config, env | False | Set to true to collect objects tags as metric labels |
VSPHERE_LIMITED_FETCH_ALARMS |
config, env | False | Fetch objects triggered alarms, and in case of hosts hdw alarms as well |
VSPHERE_LIMITED_COLLECT_HOSTS |
config, env | True | Set to false to disable collection of host metrics |
VSPHERE_LIMITED_COLLECT_DATASTORES |
config, env | True | Set to false to disable collection of datastore metrics |
VSPHERE_LIMITED_COLLECT_VMS |
config, env | True | Set to false to disable collection of virtual machine metrics |
VSPHERE_LIMITED_COLLECT_VMGUESTS |
config, env | True | Set to false to disable collection of virtual machine guest metrics |
VSPHERE_LIMITED_COLLECT_SNAPSHOTS |
config, env | True | Set to false to disable collection of snapshot metrics |
You need to set at least VSPHERE_SECTIONNAME_USER
for the section to be detected.
You can use the following parameters in the Prometheus configuration file. The params
section is used to manage multiple login/passwords.
- job_name: 'vmware_vcenter'
metrics_path: '/metrics'
static_configs:
- targets:
- 'vcenter.company.com'
relabel_configs:
- source_labels: [__address__]
target_label: __param_target
- source_labels: [__param_target]
target_label: instance
- target_label: __address__
replacement: localhost:9272
- job_name: 'vmware_esx'
metrics_path: '/metrics'
file_sd_configs:
- files:
- /etc/prometheus/esx.yml
params:
section: [esx]
relabel_configs:
- source_labels: [__address__]
target_label: __param_target
- source_labels: [__param_target]
target_label: instance
- target_label: __address__
replacement: localhost:9272
# Example of Multiple vCenter usage per #23
- job_name: vmware_export
metrics_path: /metrics
static_configs:
- targets:
- vcenter01
- vcenter02
- vcenter03
relabel_configs:
- source_labels: [__address__]
target_label: __param_target
- source_labels: [__param_target]
target_label: instance
- target_label: __address__
replacement: exporter_ip:9272
- vCenter and vSphere 6.0/6.5 have been tested.
- VM information, Snapshot, Host and Datastore basic information is exported, i.e:
# HELP vmware_snapshots VMware current number of existing snapshots
# TYPE vmware_snapshot_count gauge
vmware_snapshot_timestamp_seconds{vm_name="My Super Virtual Machine"} 2.0
# HELP vmware_snapshot_timestamp_seconds VMware Snapshot creation time in seconds
# TYPE vmware_snapshot_timestamp_seconds gauge
vmware_snapshot_age{vm_name="My Super Virtual Machine",vm_snapshot_name="Very old snaphot"} 1478146956.96092
vmware_snapshot_age{vm_name="My Super Virtual Machine",vm_snapshot_name="Old snapshot"} 1478470046.975632
# HELP vmware_datastore_capacity_size VMware Datastore capacity in bytes
# TYPE vmware_datastore_capacity_size gauge
vmware_datastore_capacity_size{ds_name="ESX1-LOCAL"} 67377299456.0
# HELP vmware_datastore_freespace_size VMware Datastore freespace in bytes
# TYPE vmware_datastore_freespace_size gauge
vmware_datastore_freespace_size{ds_name="ESX1-LOCAL"} 66349694976.0
# HELP vmware_datastore_uncommited_size VMware Datastore uncommitted in bytes
# TYPE vmware_datastore_uncommited_size gauge
vmware_datastore_uncommited_size{ds_name="ESX1-LOCAL"} 0.0
# HELP vmware_datastore_provisoned_size VMware Datastore provisoned in bytes
# TYPE vmware_datastore_provisoned_size gauge
vmware_datastore_provisoned_size{ds_name="ESX1-LOCAL"} 1027604480.0
# HELP vmware_datastore_hosts VMware Hosts number using this datastore
# TYPE vmware_datastore_hosts gauge
vmware_datastore_hosts{ds_name="ESX1-LOCAL"} 1.0
# HELP vmware_datastore_vms VMware Virtual Machines number using this datastore
# TYPE vmware_datastore_vms gauge
vmware_datastore_vms{ds_name="ESX1-LOCAL"} 0.0
# HELP vmware_host_power_state VMware Host Power state (On / Off)
# TYPE vmware_host_power_state gauge
vmware_host_power_state{host_name="esx1.company.com"} 1.0
# HELP vmware_host_cpu_usage VMware Host CPU usage in MHz
# TYPE vmware_host_cpu_usage gauge
vmware_host_cpu_usage{host_name="esx1.company.com"} 2959.0
# HELP vmware_host_cpu_max VMware Host CPU max availability in MHz
# TYPE vmware_host_cpu_max gauge
vmware_host_cpu_max{host_name="esx1.company.com"} 28728.0
# HELP vmware_host_memory_usage VMware Host Memory usage in Mbytes
# TYPE vmware_host_memory_usage gauge
vmware_host_memory_usage{host_name="esx1.company.com"} 107164.0
# HELP vmware_host_memory_max VMware Host Memory Max availability in Mbytes
# TYPE vmware_host_memory_max gauge
vmware_host_memory_max{host_name="esx1.company.com"} 131059.01953125
The VMware exporter uses theses libraries:
- pyVmomi for VMware connection
- Prometheus client_python for Prometheus supervision
- Twisted for HTTP server
The initial code is mainly inspired by:
- https://www.robustperception.io/writing-a-jenkins-exporter-in-python/
- https://github.com/vmware/pyvmomi-community-samples
- https://github.com/jbidinger/pyvmomi-tools
Forked from https://github.com/rverchere/vmware_exporter. I removed the fork so that I could do searching and everything.
Daniel Pryor pryorda
See LICENSE file