create unique task_id with timestamp #223
Conversation
Signed-off-by: youliang <tan_you_liang@hotmail.com>
| if (use_timestamp_for_task_id) | ||
| { | ||
| task_id += std::to_string( | ||
| std::chrono::duration_cast<std::chrono::seconds>( |
There was a problem hiding this comment.
Actually, why do we need the duration_cast? time_since_epoch().count() would directly return the nanoseconds since the start.
Also, should we use node time here? (node->get_clock()->now())
Signed-off-by: youliang <tan_you_liang@hotmail.com>
| if (use_timestamp_for_task_id) | ||
| { | ||
| task_id += std::to_string( | ||
| static_cast<int>(node->get_clock()->now().nanoseconds()/1e9)); |
There was a problem hiding this comment.
Wondering if we should leave the timestamp in nanoseconds and not divide by 1e9 to get seconds? If somehow two or more tasks get processed within the same second, they might have the same task_id?
There was a problem hiding this comment.
Yea, that is valid concern that i thought about too. The only concern (probably insignificant) is having nanosec as timestamp is way too long. This will result in a long task_id that might look quite ugly on the dashboard? Example shown here:
Let me nego abit, how about milisec? Since unix_millis time is regularly used when defining our rmf time field.
There was a problem hiding this comment.
Okay let's go with millis for now.
I do this this feature will be redundant once we implement a proper backup and restore functionality for this node. That way when the node is restored after a crash, it will know the last issued task_id.
Signed-off-by: youliang <tan_you_liang@hotmail.com>
|
Thanks to the approval and addressing the failed CI #225 . Will merge this in since the failed CI is not relevant to this pr |
* create unique task_id with timestamp Signed-off-by: youliang <tan_you_liang@hotmail.com> * use node time and update changelog Signed-off-by: youliang <tan_you_liang@hotmail.com> * change to milli unix sec Signed-off-by: youliang <tan_you_liang@hotmail.com> Signed-off-by: youliang <tan_you_liang@hotmail.com> Signed-off-by: Yadunund <yadunund@gmail.com>
The current way of generating a task id uses a
task_counter, which generates the task_id by appending an integertask countervalue (eg.patrol.dispatch-1). However, this poses a problem when we require the task_id to be unique even after restarting thetask dispatchernode (during deployment setup especially). Therefore, this solution generates a unique id with a timestamp. This can be enabled via ros2 param, by settinguse_timestamp_for_task_idastrue.p/s:
use_timestamp_for_task_idis false by default (using task counter)To test
Run dispatcher with use_timestamp_for_task_id as true
Try dispatch a dummy task