Cannot cancel the job

[retifrav]

List of current transfers:

http://IP-ADDRESS:PORT/core/stats

{
    ...
    "transferring": [
        {
            "bytes": 212856832,
            "eta": 93,
            "group": "job/95180",
            "name": "2019-09-26-raspbian-buster-lite.zip",
            "percentage": 46,
            "size": 454279954,
            "speed": 2750878.5798009364,
            "speedAvg": 2595620.440967448
        },
        {
            "bytes": 191885312,
            "eta": 97,
            "group": "job/95191",
            "name": "some-file.mkv",
            "percentage": 42,
            "size": 446388267,
            "speed": 2537641.0826345356,
            "speedAvg": 2609023.91652754
        }
    ],
    ...
}

Want to cancel some-file.mkv, so the jobid is 95191:

http://IP-ADDRESS:PORT/job/stop
{"jobid":"95191"}

Request is executed successfully, but the job is not cancelled.

I can also check its status:

http://IP-ADDRESS:PORT/job/status jobid="95191"

{
    "duration": 0,
    "endTime": "0001-01-01T00:00:00Z",
    "error": "",
    "finished": false,
    "group": "job/95191",
    "id": 95191,
    "output": null,
    "startTime": "2020-03-25T21:09:49.825654778Z",
    "success": false
}

What's weird is that on localhost (Windows 10 x64) it works fine, so the job gets cancelled. But when I do that on a remote host (Linux ARM), it does not.

[retifrav]

Asked the question at the rclone forum.

[retifrav]

Some backends do not support "context propagation": rclone/rclone#3257
So we can only wait for this to be implemented for all the backends.

[retifrav]

Looks like it has been fixed! I could successfully cancel an SFTP transfer with rclone v1.58, and most likely the fix was already available even in earlier versions (since v1.55, as it seems), as the upstream issue has been closed on 03.02.2021.