It also seems like the `buffer` plugin isn't addin...
# fluent-operator
j
It also seems like the
buffer
plugin isn't adding the required
@type
attribute to the matcher that it generates.
Copy code
<match **>
    @id  ClusterFluentdConfig-cluster-cluster-fluentd-config::cluster::clusteroutput::datadog-0
    <buffer>
      @type  memory
      chunk_limit_size  20MB
      flush_interval  3s
      path  /buffers/fluentd/log
      retry_forever  true
      timekey  5s
      timekey_wait  2s
    </buffer>
  </match>
Copy code
fluentd-0 fluentd 2023-07-24 18:34:09 +0000 [error]: config error file="/fluentd/etc/fluent.conf" error_class=Fluent::ConfigError error="Missing '@type' parameter on <match> directive"
@Benjamin Huo any idea here?
For refrence, here is the manifest:
Copy code
apiVersion: <http://fluentd.fluent.io/v1alpha1|fluentd.fluent.io/v1alpha1>
kind: ClusterOutput
metadata:
  name: datadog
  labels:
    <http://output.fluentd.fluent.io/scope|output.fluentd.fluent.io/scope>: "cluster"
    <http://output.fluentd.fluent.io/enabled|output.fluentd.fluent.io/enabled>: "true"
spec:
  outputs:
    - buffer:
        id: test
        type: memory
        # chunkLimitRecords: "500MB"
        chunkLimitSize: 20MB
        flushInterval: 3s
        retryForever: true
        timekey: 5s
        timekeyWait: 2s
Is anyone here using the buffer output plugin that can help test this with me?