mm::node::OsCommandConfig Struct Reference

Timing and cancellation for SomanetDrive::runOsCommand. More...

#include <somanet_drive.h>

Public Attributes
std::chrono::milliseconds	timeout {1000}
	Ceiling on the whole command.
std::chrono::milliseconds	pollInterval {10}
	Delay between response polls.
std::chrono::milliseconds	abortTimeout {10000}
	Time allowed to confirm a forced abort.
std::stop_token	stop {}
	Requesting a stop aborts the command; default never stops.

Detailed Description

Timing and cancellation for SomanetDrive::runOsCommand.

timeout is a ceiling on the whole command, not a liveness check — the drive fails a command no downstream service acknowledges on its own, after about 20 s (see OsCommandError::kTimeout) — so size it for the command being run (milliseconds for a register read, tens of seconds for a measurement). Hitting it, or a stop request on stop, makes the master abort the running command; abortTimeout then bounds how long the drive is given to report that abort.

A ceiling below the drive's own is a decision, not a safety margin: it replaces whatever the drive was going to say with "this master gave up". For a command whose failure mode is a service that will never answer, waiting past 20 s is what turns a bare timeout into the drive's own verdict.

Member Data Documentation

abortTimeout

std::chrono::milliseconds mm::node::OsCommandConfig::abortTimeout {10000}

Time allowed to confirm a forced abort.

pollInterval

std::chrono::milliseconds mm::node::OsCommandConfig::pollInterval {10}

Delay between response polls.

stop

std::stop_token mm::node::OsCommandConfig::stop {}

Requesting a stop aborts the command; default never stops.

timeout

std::chrono::milliseconds mm::node::OsCommandConfig::timeout {1000}

Ceiling on the whole command.

---

The documentation for this struct was generated from the following file:

• libs/node/somanet_drive.h