Getting Started with Device Profiles
This page is not yet available in Spanish. We are working on its translation.
If you have any questions or feedback about our current translation project,
feel free to reach out to us!The Device Onboarding Experience is not supported on this site.
Join the Preview!
Device Onboarding is in Preview. Use this form to request access.
Request AccessOverview
Device profiles define which metrics to collect and how to transform them into Datadog metrics. Each profile is expected to monitor a class of similar devices from the same vendor.
The Device Onboarding Experience provides a guided, GUI-based experience to:
- Seamlessly create and manage device profiles
- Specify tags and metrics to be collected from your network devices
- Verify the matching devices to each profile
For more information on advanced profile details, review the Profile Format Reference page.
Prerequisites
The minimum Agent version required is 7.50
or higher.
Device profile onboarding
Profile home page
The profile home page is where you can see a snapshot of the device profiles you created using the Device Onboarding Experience.
Profile details
- Build your own NDM profile by navigating to Infrastructure > Network Devices > Configuration.
- Click on SNMP Profiles > Create New Profile.
- Provide your device profile a name, vendor information(optional) and description (optional).
- Select the
SysObjectID
. This is what is used to match network devices to the device profiles that define what is collected and monitored from each device.
Add global tags for more advanced and granular options, which allows you to assign a weight to a specific metric.
Advanced Options
Modification | Description |
---|
No Modification | The device’s returned value will be used directly as the tag value. |
Format | This can be mac_address or ip_address. |
Extract Value | A regular expression used to extract the tag value from the SNMP value provided by the device. |
Mapping | This is described here. |
Scalar metrics
See Metrics Advanced Options.
Tabular metrics
See Metrics Advanced Options.
Adding tags to tabular metrics is similar to adding global tags, with two additional options:
Select whether the tag value originates from an OID
value or a segment of the table index. If Index
is chosen as the source, an index position must be specified, which then becomes the tag.
Example of using an Index position
Consider a table at OID 1.2.3.1.1
with two indices. Each row in this table includes a two-number index. Suppose column 3 of a row has OID 1.2.3.1.1.3.55.12
- here, 1.2.3.1.1
represents the table, .3
is the column number within the table, and .55.12
is the index of this specific row (all other columns for this row will also end with .55.12
). If you establish a tag with the Source set to Index
and Index Position
set to 1, the tag’s value for metrics from this table row will be 55
; if you set the index position to 2, it will be 12. If you use an index less than 1 or more than the number of indices in the table, the tag will not be populated.
See Using an Index for more information.
Use Index Transformation when you need to tag a table metric value with a value from a different table that employs a subset of this table’s index. This is not a typical scenario. You configure this by adding one or more transformation segments, each with a start and end number. These numbers index into the original table’s index to create a new index value for the new table.
Example of using Index Transformation
Consider the CPI-UNITY-MIB
module. It has a table
, cpiPduTable
, with details about a specific PDU, and another table, cpiPduBranchTable
, with information about specific PDU branches. The index of the main table is the PDU’s MAC address, such as 6.0.36.155.53.3.246
. The branch table’s index is a branch ID
followed by the PDU MAC
, therefore a branch table row index might be 1.6.0.36.155.53.3.246
.
If you want to report the current on a PDU branch, you could add cpiPduBranchCurrent
(OID 1.3.6.1.4.1.30932.1.10.1.3.110.1.3
, from the branch table) as a tabular metric. To tag this metric with the PDU name, add cpiPduName
as a tag (OID 1.3.6.1.4.1.30932.1.10.1.2.10.1.3
, from the main table), then add an index transform with start:1
and end:7
. This means the branch current metric from the branch table indexed with 1.6.0.36.155.53.3.246
would be tagged using the name from the main table indexed with 6.0.36.155.53.3.246
.
For more information see Using a column from a different table with different indexes.
Metrics Advanced Options
The advanced options for scalar and tabular metrics are the same:
- Metric Type
- One of
gauge
, rate
, monotonic_count
, or monotonic_count_and_rate
. - Scale Factor
- Before being transmitted to Datadog, the extracted value is multiplied by this factor.
- Extract Value
- This is the same as the advanced options for global tags.
Apply a profile to created devices
- After you save a profile, navigate back to the profile home page and select the Download All Profiles option. This allows you to download the
.zip
bundle which contains the yaml
files for the profiles you created. - Place the
yaml
files in the profile directory on each of the relevant installed Agents. - Restart the Datadog Agent.
- To ensure the profiles you created are accurate, confirm that NDM is receiving metrics from the matched devices as expected.
Further Reading
Más enlaces, artículos y documentación útiles: