A plugin based on Prometheus Windows Exporter to collect system / host level metrics
Prometheus Windows Exporter is a popular way to collect system level metrics from microsoft windows, such as CPU / Disk / Network / Process statistics. Fluent Bit 1.9.0 includes windows exporter metrics plugin that builds off the Prometheus design to collect system level metrics without having to manage two separate processes or agents.
The initial release of Windows Exporter Metrics contains a single collector available from Prometheus Windows Exporter and we plan to expand it over time.
Important note: Metrics collected with Windows Exporter Metrics flow through a separate pipeline from logs and current filters do not operate on top of metrics.
Key | Description | Default |
---|---|---|
The following table describes the available collectors as part of this plugin. All of them are enabled by default and respects the original metrics name, descriptions, and types from Prometheus Windows Exporter, so you can use your current dashboards without any compatibility problem.
note: the Version column specifies the Fluent Bit version where the collector is available.
This input always runs in its own thread.
In the following configuration file, the input plugin _windows_exporter_metrics collects _metrics every 2 seconds and exposes them through our Prometheus Exporter output plugin on HTTP/TCP port 2021.
You can test the expose of the metrics by using curl:
Windows service collector will retrieve all of the service information for the local node or container. we.service.where
, we.service.include
, and we.service.exclude
can be used to filter the service metrics.
To filter these metrics, users should specify a WHERE clause. This syntax is defined in the WMI Query Language(WQL).
Here is how these parameters should work:
we.service.where
is handled as a raw WHERE clause. For example, when a user specifies the parameter as follows:
This creates a WMI query like so:
The WMI mechanism will then handle it and return the information which has a "not OK" status in this example.
When defined, the we.service.include
is interpreted into a WHERE clause. If multiple key-value pairs are specified, the values will be concatenated with OR
. Also, if the values contain %
character then a LIKE
operator will be used in the clause instead of the =
operator. When a user specifies the parameter as follows:
The parameter will be interpreted as:
The WMI query will be called with the translated parameter as:
When defined, the we.service.exclude
is interpreted into a WHERE clause. If multiple key-value pairs are specified, the values will be concatenated with AND
.
Also, if the values contain %
character then a LIKE
operator will be used in the translated clause instead of the !=
operator. When a user specifies the parameter as follows:
The parameter will be interpreted as:
The WMI query will be called with the translated parameter as:
we.service.where
, we.service.include
, and we.service.exclude
can all be used at the same time subject to the following rules.
we.service.include
translated and applied into the where clause in the service collector
we.service.exclude
translated and applied into the where clause in the service collector
If the we.service.include
is applied, translated we.service.include
and we.service.exclude
conditions are concatenated with AND
.
we.service.where
is just handled as-is into the where clause in the service collector .
If either of the above parameters is applied, the clause will be applied with AND (
the value of we.service.where
)
.
For example, when a user specifies the parameter as follows:
The WMI query will be called with the translated parameter as:
Our current plugin implements a sub-set of the available collectors in the original Prometheus Windows Exporter, if you would like that we prioritize a specific collector please open a Github issue by using the following template: - in_windows_exporter_metrics
Name | Description | OS | Version |
---|---|---|---|
scrape_interval
The rate at which metrics are collected from the host operating system
5 seconds
we.logical_disk.allow_disk_regex
Specify the regex for logical disk metrics to allow collection of. Collect all by default.
"/.+/"
we.logical_disk.deny_disk_regex
Specify the regex for logical disk metrics to prevent collection of/ignore. Allow all by default.
NULL
we.net.allow_nic_regex
Specify the regex for network metrics captured by the name of the NIC, by default captures all NICs but to exclude adjust the regex.
"/.+/"
we.service.where
Specify the WHERE clause for retrieving service metrics.
NULL
we.service.include
Specify the key value pairs for the include condition for the WHERE clause of service metrics.
NULL
we.service.exclude
Specify the key value pairs for the exclude condition for the WHERE clause of service metrics.
NULL
we.process.allow_process_regex
Specify the regex covering the process metrics to collect. Collect all by default.
"/.+/"
we.process.deny_process_regex
Specify the regex for process metrics to prevent collection of/ignore. Allow all by default.
NULL
collector.cpu.scrape_interval
The rate in seconds at which cpu metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.net.scrape_interval
The rate in seconds at which net metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.logical_disk.scrape_interval
The rate in seconds at which logical_disk metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.cs.scrape_interval
The rate in seconds at which cs metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.os.scrape_interval
The rate in seconds at which os metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.thermalzone.scrape_interval
The rate in seconds at which thermalzone metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.cpu_info.scrape_interval
The rate in seconds at which cpu_info metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.logon.scrape_interval
The rate in seconds at which logon metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.system.scrape_interval
The rate in seconds at which system metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.service.scrape_interval
The rate in seconds at which service metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.memory.scrape_interval
The rate in seconds at which memory metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.paging_file.scrape_interval
The rate in seconds at which paging_file metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
collector.process.scrape_interval
The rate in seconds at which process metrics are collected from the host operating system. If a value greater than 0 is used then it overrides the global default otherwise the global default is used.
0 seconds
metrics
To specify which metrics are collected from the host operating system.
"cpu,cpu_info,os,net,logical_disk,cs,thermalzone,logon,system,service"
cpu
Exposes CPU statistics.
Windows
v1.9
net
Exposes Network statistics.
Windows
v2.0.8
logical_disk
Exposes logical_disk statistics.
Windows
v2.0.8
cs
Exposes cs statistics.
Windows
v2.0.8
os
Exposes OS statistics.
Windows
v2.0.8
thermalzone
Exposes thermalzone statistics.
Windows
v2.0.8
cpu_info
Exposes cpu_info statistics.
Windows
v2.0.8
logon
Exposes logon statistics.
Windows
v2.0.8
system
Exposes system statistics.
Windows
v2.0.8
service
Exposes service statistics.
Windows
v2.1.6
memory
Exposes memory statistics.
Windows
v2.1.9
paging_file
Exposes paging_file statistics.
Windows
v2.1.9
process
Exposes process statistics.
Windows
v2.1.9