nicolargo/glances · error · HTTPException

Unknown PID process {pid}

Error message

Unknown PID process {pid}

What it means

HTTP 404 raised by PUT /api/4/processes/{pid}/extended (set extended process) when glances_processes.get_stats(int(pid)) returns a falsy value, meaning no stats are currently cached for that PID. The process may have exited between listing and selection, or it is not visible to Glances' process collector.

Source

Thrown at glances/outputs/glances_restful_api.py:1429

        try:
            args_json = self._sanitize_args()[item]
        except Exception as e:
            raise HTTPException(status.HTTP_404_NOT_FOUND, f"Cannot get args item ({str(e)})")

        return GlancesJSONResponse(args_json)

    def _api_set_extended_processes(self, pid: str):
        """Glances API RESTful implementation.

        Set the extended process stats for the given PID
        HTTP/200 if OK
        HTTP/400 if PID is not found
        HTTP/404 if others error
        """
        process_stats = glances_processes.get_stats(int(pid))

        if not process_stats:
            raise HTTPException(status.HTTP_404_NOT_FOUND, f"Unknown PID process {pid}")

        glances_processes.extended_process = process_stats

        return GlancesJSONResponse(True)

    def _api_disable_extended_processes(self):
        """Glances API RESTful implementation.

        Disable extended process stats
        HTTP/200 if OK
        HTTP/400 if PID is not found
        HTTP/404 if others error
        """
        glances_processes.extended_process = None

        return GlancesJSONResponse(True)

    def _api_get_extended_processes(self):

View on GitHub (pinned to a240d8dfb3)

Solutions

1. Retry shortly — the process cache refreshes on the stats update cycle, newly spawned PIDs appear then
2. Verify with GET /api/4/processes that the PID is in the current list before extending it
3. Check --enable-process-extended is supported and no process filter (name/user regex) excludes the target
4. If monitoring containers, query the PID from the same PID namespace as glances

Example fix

# before
PUT /api/4/processes/4242/extended
# after (confirm existence first)
GET /api/4/processes | jq '.[] | select(.pid==4242)'
PUT /api/4/processes/4242/extended

live = {p['pid'] for p in requests.get(f'{base}/api/4/processes').json()}
if pid not in live:
    raise LookupError(f'PID {pid} not visible to glances')

def pid_exists(pid: int) -> bool:
    return any(p['pid'] == pid for p in requests.get(f'{base}/api/4/processes').json())

try:
    r = requests.put(f'{base}/api/4/processes/{pid}/extended')
    r.raise_for_status()
except requests.HTTPError as e:
    if e.response.status_code == 404:
        # process gone: stop tracking
        tracked.discard(pid)
    else:
        raise

When it happens

Trigger: PUT /api/4/processes/1234/extended where PID 1234 is dead, not yet in the process cache (stats update every refresh interval), hidden by process filtering, or inside another PID namespace (container).

Common situations: Web UI race: user clicks a process that exits before the request; monitoring containerized workloads from the host (or vice versa); strict process filtering excluding the PID.

Related errors

• Cannot get help view data ({str(e)})
• Cannot get plugin list ({str(e)})
• Cannot get stats ({str(e)})
• Cannot get limits ({str(e)})
• Cannot get views ({str(e)})

AI-assisted analysis of nicolargo/glances@a240d8dfb3 (2026-08-27). Data as JSON: /api/errors/d9d91bcd6499dce6. Report an issue: GitHub.