addK8sMeta
Used to get from some fields in the event (such as the path of the log file).
Source interceptor.
pod.uid
namespace
andpod.name
container.id
With any one of the above three kinds of index information, Loggie can query the specific Pod according to the index, and add additional kubernetes meta-information such as ${node.name}
, ${namespace}
, ${pod.uid}
, ${pod.name}
, etc. to the event for subsequent analysis and processing.
Example
interceptors:
- type: addK8sMeta
pattern: "/var/log/${pod.uid}/${pod.name}/"
addFields:
nodename: "${node.name}"
namespace: "${namespace}"
podname: "${pod.name}"
pattern
字段 | 类型 | 是否必填 | 默认值 | 含义 |
---|---|---|---|---|
pattern | string | true | Matching model for extracting fields |
Must contain one of:
pod.uid
namespace
andpod.name
container.id
For example: /var/log/${pod.uid}/${pod.name}/
patternFields
字段 | 类型 | 是否必填 | 默认值 | 含义 |
---|---|---|---|---|
patternFields | string | false | By default, the filename in the system field will be obtained from the event. In this case, you need to use the file source | Fields for the extracting pattern from the event |
fieldsName
字段 | 类型 | 是否必填 | 默认值 | 含义 |
---|---|---|---|---|
fieldsName | string | false | kubernetes | Fields to add meta information |
addFields
字段 | 类型 | 是否必填 | 默认值 | 含义 |
---|---|---|---|---|
addFields | map | false | Meta information to be added |
The currently supported meta-information fields are:
${cluster}
: Cluster information.discovery.kubernetes.cluster
in system configuration.${node.name}
${namespace}
${workload.kind}
: Deployment/StatefulSet/DaemonSet/Job, etc.${workload.name}
: the name of the workload${pod.uid}
${pod.name}