Track your Tasks

Track your Tasks#

The Inductiva API allows you to run multiple simulations remotely, and for every simulation you submit to the API, you generate a task. The Inductiva CLI makes it easy to monitor and manage all tasks you generated with the list subcommand that displays details about the most recent tasks you launched. You can also track a specific task using its unique--id.

See also

Learn more about how tasks are generated through the Inductiva API.

Here’s an example of how you can use these features:

# List the last four tasks with their details
$ inductiva tasks list -n 4
       ID                              SIMULATOR          STATUS         SUBMITTED              STARTED                COMPUTATION TIME         RESOURCE TYPE
       jxwt0rm8s8xspdfcegtgkkana       splishsplash       started        08 Feb, 13:25:49       08 Feb, 13:26:04       *0:00:05                 c2-standard-4
       n0zcac8rmw7xydbis3m407kb4       splishsplash       started        08 Feb, 13:25:48       08 Feb, 13:26:03       *0:00:07                 c2-standard-4
       8nmpn4h99nyfpo4da9jw2405q       splishsplash       started        08 Feb, 13:25:47       08 Feb, 13:26:02       *0:00:09                 c2-standard-4
       so6i93pi74b89rndircubp3v2       splishsplash       started        08 Feb, 13:25:47       08 Feb, 13:26:02       *0:00:10                 c2-standard-4

To continuously monitor task progress, you can employ a watch method on Linux or Mac, refreshing task information at set intervals.

In this example, we refreshed task information every 10 seconds:

# Monitor task status updates every 10 seconds
$ watch -n 10 inductiva tasks list --id jxwt0rm8s8xspdfcegtgkkana
Every 10.0s: inductiva tasks list --id jxwt0rm8s8xspdfcegtgkkana                                                                                 


       ID                              SIMULATOR          STATUS         SUBMITTED              STARTED                COMPUTATION TIME         RESOURCE TYPE
       jxwt0rm8s8xspdfcegtgkkana       splishsplash       success        08 Feb, 13:25:49       08 Feb, 13:26:04       0:00:35                  c2-standard-4

If you identify an error in your simulation or simply wish to stop it, the kill subcommand allows you to terminate it, thus freeing up the computational resources for other tasks:

# Terminate a specific task
$ inductiva tasks kill cmvsc9qhz5iy86f6pef8uyxqt
You are about to kill the following tasks:
  - cmvsc9qhz5iy86f6pef8uyxqt 
Are you sure you want to proceed (y/[N])? y
Successfully sent kill request for task cmvsc9qhz5iy86f6pef8uyxqt.

Following the termination, you can track the task status by calling task.get_status(), or via the Inductiva CLI:

# Check the status of the terminated task
$ inductiva tasks list --id cmvsc9qhz5iy86f6pef8uyxqt

      ID                              SIMULATOR          STATUS         SUBMITTED              STARTED         COMPUTATION TIME         RESOURCE TYPE
       cmvsc9qhz5iy86f6pef8uyxqt       splishsplash       killed         08 Feb, 13:41:06       n/a             n/a                      n/a

If you want to wait for confirmation when you send the request to kill the task, you can use the --wait-timeout flag with the kill subcommand specifying the number of seconds to wait for confirmation:

# Request to kill a task with a timeout for confirmation
$ inductiva tasks kill cmvsc9qhz5iy86f6pef8uyxqt --wait-timeout 10

Finally, the Inductiva CLI allows you to download tasks using:

$ inductiva tasks download task_1_id task_2_id

Additionally, we can specify specific files to download using --filenames and we can specify the output directory with --output_dir. When --output_dir is provided the tasks will be download under output_dir/tasks_id.