Log upstream provider rejections at warn/error level (#27238)

When an upstream provider rejects a request (e.g. a 400 for a
max_tokens value above the model's ceiling), the actionable error
message was only published to event sinks, which are invisible unless
an event function or webhook is configured. Admins had to query the
provider's API directly to diagnose failures (open-webui#27237).

Add a single log line in publish_model_provider_request_failed — the
chokepoint every upstream failure path (OpenAI-compatible chat,
embeddings, responses, token counting, and Ollama) already routes
through — recording status, provider, url, model, error code, and the
upstream message truncated to 1000 chars. 4xx logs at WARNING, 5xx at
ERROR. Client-facing responses are unchanged, so no additional error
detail is exposed in the chat.


Claude-Session: https://claude.ai/code/session_018VecyiPejru1EVF5yfe2sU

Co-authored-by: Claude <noreply@anthropic.com>

backend/open_webui/events.py

@@ -1193,6 +1193,20 @@ async def publish_model_provider_request_failed(
         else 'upstream_error'
     )
 
+    # Server-log only; the upstream error body is otherwise invisible to admins
+    # (event sinks require an event function or webhook to be configured).
+    log.log(
+        logging.ERROR if status >= 500 else logging.WARNING,
+        'Upstream %s request failed: HTTP %d (%s) url=%s model=%s code=%s message=%s',
+        provider,
+        status,
+        error_type,
+        base_url,
+        requested_model or '-',
+        error_code or '-',
+        error_text[:MAX_STRING_LENGTH] or '-',
+    )
+
     data = {
         'error_type': error_type,
         'status': status,