This release substantially changes the behavior of the exporter and telegraf containers in the pod: previously, a failure in either of these processes would lead to the container failing and being restarted by the kubelet, potentially leading to the entire pod being marked as unavailable. Now by default an error is logged and the process is restarted by the wrapper script, but the container does not exit.
An exitOnError helm option has been added to both the exporter and telegraf config value sections: set each to "true" to restore the old behavior.
This release substantially changes the behavior of the exporter and telegraf containers in the pod: previously, a failure in either of these processes would lead to the container failing and being restarted by the kubelet, potentially leading to the entire pod being marked as unavailable. Now by default an error is logged and the process is restarted by the wrapper script, but the container does not exit.
An
exitOnErrorhelm option has been added to both theexporterandtelegrafconfig value sections: set each to "true" to restore the old behavior.