boost::redis::config::health_check_interval

Time span between successive health checks. Set to zero to disable health‐checks.

Synopsis

Declared in <boost/redis/config.hpp>

std::chrono::steady_clock::duration health_check_interval = std::chrono::seconds{2};

Description

When this value is set to a non‐zero duration, basic_connection::async_run will issue PING commands whenever no command is sent to the server for more than health_check_interval. You can configure the message passed to the PING command using health_check_id.

Enabling health checks also sets timeouts to individual network operations. The connection is considered dead if:

  • No byte can be written to the server after `health_check_interval`.

  • No byte is read from the server after `2 * health_check_interval`.

If the health checker finds that the connection is unresponsive, it will be closed, and a reconnection will be triggered, as if a network error had occurred.

The exact timeout values are not part of the interface, and might change in future versions.

When using Sentinel, this setting applies to masters and replicas. Sentinels are not health‐checked.

Created with MrDocs