MrSphay/odysseus
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(models): stabilize proxy endpoint refresh behavior

Browse Source 
* fix: support large proxy model endpoint refresh

Large OpenAI-compatible proxy endpoints can expose hundreds of models and make /v1/models slow. Treating those endpoints like local model servers caused model picker opens and background probes to repeatedly hit /models, producing timeouts and making otherwise usable endpoints appear offline.

Make model endpoint discovery cached-first for normal UI usage, add explicit proxy/API classification and refresh policy fields, exclude proxy/API endpoints from aggressive local probing, and preserve cached models when refresh fails.

Manual Test/Add/Refresh actions still fetch the full model list with longer timeouts so users can intentionally import large proxy model lists without blocking normal model picker usage.

* fix: preserve endpoint ping status semantics
This commit is contained in:
Yuri
2026-06-04 00:56:11 -03:00
committed by GitHub
parent eee2167502
commit a2e691da2b
10 changed files with 1323 additions and 231 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
core/database.py
View File

@@ -342,6 +342,14 @@ class ModelEndpoint(TimestampMixin, Base):
cached_models = Column(Text, nullable=True) # JSON list of last-known model IDs (avoids probe on list)
pinned_models = Column(Text, nullable=True) # JSON list of admin-pinned model IDs (manual, may not appear in /v1/models)
model_type = Column(String, nullable=True, default="llm") # "llm" or "image"
# auto = classify by URL; local = self-hosted server; api/proxy = external
# OpenAI-compatible API even when reachable through a private/tailnet IP.
endpoint_kind = Column(String, nullable=True, default="auto")
# auto = background refresh with TTL/backoff; manual/disabled = cached-first
# only unless an explicit endpoint probe is requested.
model_refresh_mode = Column(String, nullable=True, default="auto")
model_refresh_interval = Column(Integer, nullable=True, default=None)
model_refresh_timeout = Column(Integer, nullable=True, default=None)
# Whether models on this endpoint accept OpenAI-style function
# schemas + emit `tool_calls`. Auto-detected at Cookbook auto-
# register time from `--enable-auto-tool-choice` in the serve cmd;
@@ -809,6 +817,29 @@ def _migrate_add_model_type_column():
except Exception as e:
logging.getLogger(__name__).warning(f"model_type migration failed: {e}")
def _migrate_add_model_endpoint_refresh_columns():
"""Add endpoint classification / refresh policy columns if missing."""
import sqlite3
db_path = DATABASE_URL.replace("sqlite:///", "")
if not os.path.exists(db_path):
return
try:
conn = sqlite3.connect(db_path)
cursor = conn.execute("PRAGMA table_info(model_endpoints)")
columns = [row[1] for row in cursor.fetchall()]
if columns and "endpoint_kind" not in columns:
conn.execute("ALTER TABLE model_endpoints ADD COLUMN endpoint_kind TEXT DEFAULT 'auto'")
if columns and "model_refresh_mode" not in columns:
conn.execute("ALTER TABLE model_endpoints ADD COLUMN model_refresh_mode TEXT DEFAULT 'auto'")
if columns and "model_refresh_interval" not in columns:
conn.execute("ALTER TABLE model_endpoints ADD COLUMN model_refresh_interval INTEGER")
if columns and "model_refresh_timeout" not in columns:
conn.execute("ALTER TABLE model_endpoints ADD COLUMN model_refresh_timeout INTEGER")
conn.commit()
conn.close()
except Exception as e:
logging.getLogger(__name__).warning(f"model_endpoints refresh-policy migration failed: {e}")
def _migrate_add_task_run_model_column():
"""Add model column to task_runs if it doesn't exist (records which model ran)."""
import sqlite3
@@ -1539,6 +1570,7 @@ def init_db():
_migrate_add_pinned_models_column()
_migrate_add_notes_sort_order()
_migrate_add_model_type_column()
_migrate_add_model_endpoint_refresh_columns()
_migrate_add_model_endpoint_owner_column()
_migrate_add_supports_tools_column()
_migrate_add_task_run_model_column()