Runtime parameters
Runtime options are cluster-wide settings that you can change on a running cluster without reinstalling Portworx Enterprise. Set runtime options by using the pxctl cluster options update --runtime-options command.
Set runtime options
The --runtime-options flag replaces the entire set of runtime options with the values that you specify. It does not merge the specified values with the existing configuration. If you omit a previously configured option, the option is removed from the runtime configuration.
To preserve existing values, review the current runtime options before making changes:
pxctl cluster options list
Include every option that you want to retain in the update command, along with the option that you are changing. For example, to add max_resync_bandwidth_mbps to a cluster that already has device_delete_after_discard set, run:
pxctl cluster options update --runtime-options "device_delete_after_discard=1,max_resync_bandwidth_mbps=850"
Successfully updated cluster-wide options
Run the following command again to verify that all expected runtime options are set:
pxctl cluster options list
Do not modify runtime options unless necessary. Contact Portworx Support before changing a runtime option.
KVDB runtime options
The following table provides the internal KVDB runtime options that can be used to customize the KVDB database during its runtime.
| Parameter | Description | Default value |
|---|---|---|
| kvdb- heartbeat- interval | Specifies the frequency in milliseconds at which each KVDB node will send heartbeats to its peers. | 250 ms |
| kvdb- election- timeout | Specifies the timeout for leader election in milliseconds. | 1000 ms. By default, Portworx will set the election timeout to 5 times the heartbeat interval. |
| kvdb- snap- count | Specifies the number of committed transactions to trigger a snapshot to disk. | 10000 |
| kvdb- debug- mode | Specifies the debug mode for KVDB debug logs. | 1 (enabled by default) |
| kvdb- pre- vote | Runs the pre-vote election algorithm when a node rejoins. If enabled, it avoids cluster disruptions caused by leader elections when a new node joins. | 1 (enabled by default) |