Hi,
We've been using the old Cloudwatch Logs plugin for Fluent for some time, and want to start using the new one. However, we make use of templating to set the log group name and log stream name (as described here):
[OUTPUT]
name cloudwatch
match *
region eu-central-1
log_group_name $(log_group_name)
log_stream_name $(log_stream_name)
auto_create_group true
log_retention_days 30
retry_limit false
The log_group_name and log_stream_name keys are present in the log message, the values are then used to set the log group name and the log stream name. This, however, does not seem to work with the new plugin. I noticed that the new plugin should support it through the 'record accessor' syntax, but the following does not result in the log_group_name and log_stream_name being evaluated correctly:
[OUTPUT]
name cloudwatch_logs
match *
region eu-central-1
log_group_name $log_group_name
log_stream_name $log_stream_name
auto_create_group true
log_retention_days 30
retry_limit false
Am I misunderstanding the record accessor syntax? Or does this only apply to metadata, rather than the log message itself?
Hi,
We've been using the old Cloudwatch Logs plugin for Fluent for some time, and want to start using the new one. However, we make use of templating to set the log group name and log stream name (as described here):
The
log_group_nameandlog_stream_namekeys are present in the log message, the values are then used to set the log group name and the log stream name. This, however, does not seem to work with the new plugin. I noticed that the new plugin should support it through the 'record accessor' syntax, but the following does not result in the log_group_name and log_stream_name being evaluated correctly:Am I misunderstanding the record accessor syntax? Or does this only apply to metadata, rather than the log message itself?