What causes "url in downloading, %s" in XX-net/XX-Net?

Wait for the in-flight download to finish and retry the operation once

url in downloading, %s

Error message

url in downloading, %s

What it means

update_from_github.download_file keeps a global progress dict keyed by URL; if an entry exists with status 'downloading', a concurrent call for the same URL is rejected with this warning and returns False, to prevent two threads writing the same file.

Source

Thrown at code/default/launcher/update_from_github.py:92

            "host": "127.0.0.1",
            "port": 8086,
            "user": None,
            "pass": None
        }, timeout=timeout, cert=cert)

    res = client.request("GET", url, read_payload=False)
    return res


def download_file(url, filename):
    if url not in progress:
        progress[url] = {}
        progress[url]["status"] = "downloading"
        progress[url]["size"] = 1
        progress[url]["downloaded"] = 0
    else:
        if progress[url]["status"] == "downloading":
            xlog.warn("url in downloading, %s", url)
            return False

    for i in range(0, 2):
        try:
            xlog.info("download %s to %s, retry:%d", url, filename, i)
            req = request(url, i, timeout=120)
            if not req:
                continue

            start_time = time.time()
            timeout = 300

            if req.chunked:
                # don't known the file size, set to large for show the progress
                progress[url]["size"] = 20 * 1024 * 1024

                downloaded = 0
                with open(filename, 'wb') as fp:

View on GitHub (pinned to cfa5bc17b6)

Solutions

Wait for the in-flight download to finish and retry the operation once
If it never clears (stale state after a crash), restart the launcher to reset the progress dict
Avoid triggering update operations concurrently from the UI
Code-level: reset progress[url] status in a finally block so failures don't leave 'downloading' state

Example fix

// before (in update_from_github.download_file)
progress[url]["status"] = "downloading"
try:
    do_download()
finally:
    progress.pop(url, None)

// concept: always clear the per-url lock state so a crashed download does not block the next attempt

Defensive patterns

Strategy: validation

Validate before calling

from update_from_github import progress
def url_free(url):
    return progress.get(url, {}).get('status') != 'downloading'
assert url_free(url), 'download already in flight; wait or restart'

Try / catch

try:
    ok = download_file(url, path)
    if not ok and progress.get(url,{}).get('status') == 'downloading':
        wait_for_download(url)  # another thread owns it
except Exception as e:
    log.warning('download error: %r', e)
    progress.pop(url, None)

Prevention

Avoid triggering concurrent updates of the same module
Restart the launcher if a URL seems permanently 'downloading' (stale state)
Wrap download state updates in try/finally to always reset status

When it happens

Trigger: Two callers invoke download_file with the same URL concurrently — e.g. get_github_versions and download_overwrite_new_version racing, or a user double-triggering update while a download is in flight; also stale 'downloading' state left by a crashed thread.

Common situations: Double-clicking update / overlapping update checks in the web UI; a previous download thread died without resetting progress[url]['status'], permanently blocking that URL.

Related errors

AI-assisted analysis of XX-net/XX-Net@cfa5bc17b6 (2026-08-27). Data as JSON: /api/errors/058829e47d2d2d2b. Report an issue: GitHub.