fluentd_memory
Description
The fluentd_memory parameter configures the memory request and limit for the Fluentd log collector DaemonSet. Fluentd runs on every node in the cluster and forwards application and system logs. Both the Kubernetes memory request and limit are set to the same value, giving the Fluentd pod a Guaranteed QoS class.
Default Value
The default value for fluentd_memory is 200Mi.
Use Cases
- High Log Throughput: Increase memory for racks with many services, verbose application logging, or high request rates to prevent Fluentd OOM restarts and log loss.
- GPU / ML Workloads: Workloads generating large volumes of stdout/stderr may require
512Mior1Gi. - Resource Optimization: Reduce memory below
200Mion racks with low log volume to reclaim node resources.
Setting Parameters
To set the fluentd_memory parameter, use the following command:
$ convox rack params set fluentd_memory=512Mi -r rackName
Setting parameters... OK
This command sets the Fluentd memory allocation to 512Mi.
Additional Information
The value must be a valid Kubernetes memory string matching the pattern ^\d+(Mi|Gi)$ (e.g., 200Mi, 512Mi, 1Gi). When Fluentd is OOM-killed due to high log throughput, in-flight log buffers are lost. Increasing this value prevents gaps in log delivery during high-throughput periods.
The default of 200Mi matches the previously hardcoded allocation, so existing racks are unaffected.
See Also
