===== REDPOINT OLLAMA CLUSTER HEALTH LATEST ===== STAMP=2026-07-13_22:03:29 PROXY=http://127.0.0.1:11434 OPENWEBUI=http://127.0.0.1:3001 OPENWEBUI_HTTP=200 ===== VERSION ===== {"version":"redpoint-ollama-cluster-proxy-v23","backend_count":16,"mode":"config-driven-policy-aliases-prefixed-visible-workers","config_dir":"/etc/redpoint/ollama-cluster","aliases":["rp-14b:latest","rp-254-fast:latest","rp-all-qwen25:latest","rp-fast-wide:latest","rp-fast:latest","rp-qwen35-9b:latest"]} ===== HEALTH ===== {"backend_count":16,"visible_model_count":31,"aliases":{"rp-fast:latest":[["n254","qwen2.5:1.5b"],["n26","qwen2.5:1.5b"],["n27","qwen2.5:1.5b"],["n29","qwen2.5:1.5b"]],"rp-fast-wide:latest":[["n254","qwen2.5:1.5b"],["n26","qwen2.5:1.5b"],["n27","qwen2.5:1.5b"],["n29","qwen2.5:1.5b"],["n23","qwen2.5:1.5b"]],"rp-all-qwen25:latest":[["n254","qwen2.5:1.5b"],["n26","qwen2.5:1.5b"],["n27","qwen2.5:1.5b"],["n29","qwen2.5:1.5b"],["n23","qwen2.5:1.5b"],["n205","qwen2.5:1.5b"],["n21","qwen2.5:1.5b"],["n22","qwen2.5:1.5b"],["n30","qwen2.5:1.5b"],["n31","qwen2.5:1.5b"]],"rp-254-fast:latest":[["n254","qwen2.5:1.5b"]],"rp-14b:latest":[["n254","qwen3:14b"],["n209","qwen3:14b"]],"rp-qwen35-9b:latest":[["n254","qwen3.5:9b"],["n26","qwen3.5:9b"],["n27","qwen3.5:9b"],["n31","qwen3.5:9b"]]},"default_exclude":["n204/qwen3:1.7b","n254/qwen3:14b"],"config_dir":"/etc/redpoint/ollama-cluster"} ===== MODEL COUNTS ===== MODEL_COUNT=31 ALIAS_COUNT=6 ===== ALIASES ===== rp-14b:latest rp-254-fast:latest rp-all-qwen25:latest rp-fast-wide:latest rp-fast:latest rp-qwen35-9b:latest ===== KEEPWARM ===== rp-fast:latest OK backend=n26 seconds=50.55 response='OK' rp-fast-wide:latest OK backend=n29 seconds=60.7 response='OK' rp-254-fast:latest OK backend=n254 seconds=25.58 response='OK' ===== ERRORS ===== NONE ===== END ===== ===== FLEET MATRIX ===== ===== REDPOINT OLLAMA FLEET MATRIX LATEST ===== STAMP=2026-07-16_04:13:33 ACTIVE_BACKEND_COUNT=16 ACTIVE_REACHABLE=13 ACTIVE_TOTAL_MODELS=21 CANDIDATE_COUNT=4 CANDIDATE_REACHABLE=1 ===== ACTIVE BACKENDS ===== n201 UP enabled=True models=1 version=0.30.10 role=pve url=http://192.168.0.201:11434 models=qwen3:1.7b n203 UP enabled=True models=1 version=0.30.10 role=pve url=http://192.168.0.203:11434 models=qwen3:1.7b n204 UP enabled=True models=1 version=0.30.10 role=pve url=http://192.168.0.204:11434 models=qwen3:1.7b n205 UP enabled=True models=1 version=0.31.1 role=pve url=http://192.168.0.205:11434 models=qwen2.5:1.5b n206 UP enabled=True models=1 version=0.30.10 role=pve url=http://192.168.0.206:11434 models=qwen3:1.7b n207 UP enabled=True models=1 version=0.30.10 role=pve url=http://192.168.0.207:11434 models=qwen3:1.7b n209 UP enabled=True models=2 version=0.30.10 role=pve-ai url=http://127.0.0.1:11436 models=qwen3:1.7b, qwen3:14b n21 DOWN enabled=True models=0 version=None role=windows url=http://192.168.0.21:11434 error=version: URLError(TimeoutError('timed out')) n22 UP enabled=True models=1 version=0.31.2 role=windows url=http://192.168.0.22:11434 models=qwen2.5:1.5b n23 UP enabled=True models=1 version=0.24.0 role=windows url=http://192.168.0.23:11434 models=qwen2.5:1.5b n254 UP enabled=True models=5 version=0.30.6 role=windows-m2 url=http://192.168.0.254:11434 models=qwen2.5:1.5b, qwen3.5:0.8b, qwen3.5:4b, qwen3.5:9b, qwen3:14b n26 DOWN enabled=True models=0 version=None role=windows-fast url=http://192.168.0.26:11434 error=version: URLError(OSError(113, 'No route to host')) n27 DOWN enabled=True models=0 version=None role=windows-fast url=http://192.168.0.27:11434 error=version: URLError(OSError(113, 'No route to host')) n29 UP enabled=True models=1 version=0.31.1 role=windows-fast url=http://192.168.0.29:11434 models=qwen2.5:1.5b n30 UP enabled=True models=1 version=0.30.10 role=windows url=http://192.168.0.30:11434 models=qwen2.5:1.5b n31 UP enabled=True models=4 version=0.31.1 role=windows url=http://192.168.0.31:11434 models=qwen2.5:1.5b, qwen2.5:3b, qwen3.5:4b, qwen3.5:9b ===== FUTURE / OFFLINE CANDIDATES ===== n20 DOWN state=offline_or_blocked models=0 version=None url=http://192.168.0.20:11434 error=version: URLError(ConnectionRefusedError(111, 'Connection refused')) n24 DOWN state=offline_or_blocked models=0 version=None url=http://192.168.0.24:11434 error=version: URLError(OSError(113, 'No route to host')) n25 DOWN state=credential_rejected models=0 version=None url=http://192.168.0.25:11434 error=version: URLError(OSError(113, 'No route to host')) n28 UP state=offline_or_blocked models=2 version=0.30.8 url=http://192.168.0.28:11434 models=qwen3.5:0.8b, qwen3.5:4b ===== ALIASES ===== rp-fast:latest -> n254/qwen2.5:1.5b, n26/qwen2.5:1.5b, n27/qwen2.5:1.5b, n29/qwen2.5:1.5b rp-fast-wide:latest -> n254/qwen2.5:1.5b, n26/qwen2.5:1.5b, n27/qwen2.5:1.5b, n29/qwen2.5:1.5b, n23/qwen2.5:1.5b rp-all-qwen25:latest -> n254/qwen2.5:1.5b, n26/qwen2.5:1.5b, n27/qwen2.5:1.5b, n29/qwen2.5:1.5b, n23/qwen2.5:1.5b, n205/qwen2.5:1.5b, n21/qwen2.5:1.5b, n22/qwen2.5:1.5b, n30/qwen2.5:1.5b, n31/qwen2.5:1.5b rp-254-fast:latest -> n254/qwen2.5:1.5b rp-14b:latest -> n254/qwen3:14b, n209/qwen3:14b rp-qwen35-9b:latest -> n254/qwen3.5:9b, n26/qwen3.5:9b, n27/qwen3.5:9b, n31/qwen3.5:9b ===== POLICY ===== {"default_exclude":[["n204","qwen3:1.7b"],["n254","qwen3:14b"]],"tag_cache_seconds":10,"connect_timeout_seconds":5,"generation_timeout_seconds":3600} RESULT=FLEET_MATRIX_OK ===== END ===== ===== PERFORMANCE LEDGER ===== ===== REDPOINT OLLAMA PERFORMANCE LEDGER LATEST ===== STAMP=2026-07-16_03:43:29 BACKENDS_TESTED=16 FAST=4 GOOD=0 SLOW_OK=6 SLOW_EXCLUDE_DEFAULT=2 BAD_RESPONSE=0 FAIL=4 ===== BACKEND PERFORMANCE ===== n204 FAIL status=fail seconds= model=qwen3:1.7b role=pve response='' error=TimeoutError('timed out') n21 FAIL status=fail seconds= model= role=windows response='' error=URLError(TimeoutError('timed out')) n26 FAIL status=fail seconds= model= role=windows-fast response='' error=URLError(OSError(113, 'No route to host')) n27 FAIL status=fail seconds= model= role=windows-fast response='' error=URLError(OSError(113, 'No route to host')) n31 FAST status=ok seconds=0.28 model=qwen2.5:1.5b role=windows response='OK' n254 FAST status=ok seconds=0.3 model=qwen2.5:1.5b role=windows-m2 response='OK' n29 FAST status=ok seconds=0.32 model=qwen2.5:1.5b role=windows-fast response='OK' n23 FAST status=ok seconds=0.34 model=qwen2.5:1.5b role=windows response='OK' n203 SLOW_EXCLUDE_DEFAULT status=ok seconds=106.13 model=qwen3:1.7b role=pve response='' n201 SLOW_EXCLUDE_DEFAULT status=ok seconds=25.96 model=qwen3:1.7b role=pve response='' n205 SLOW_OK status=ok seconds=10.18 model=qwen2.5:1.5b role=pve response='OK' n206 SLOW_OK status=ok seconds=12.03 model=qwen3:1.7b role=pve response='' n22 SLOW_OK status=ok seconds=13.9 model=qwen2.5:1.5b role=windows response='OK' n30 SLOW_OK status=ok seconds=6.3 model=qwen2.5:1.5b role=windows response='OK' n209 SLOW_OK status=ok seconds=7.46 model=qwen3:1.7b role=pve-ai response='' n207 SLOW_OK status=ok seconds=9.0 model=qwen3:1.7b role=pve response='' ===== ROUTING HINTS ===== FAST_QWEN25_POOL=n23,n254,n29,n31 GOOD_QWEN25_POOL=n23,n254,n29,n31 SLOW_OR_FAIL=n201,n203,n204,n205,n206,n207,n209,n21,n22,n26,n27,n30 NOTE=No routing was changed by this audit. RESULT=PERFORMANCE_LEDGER_OK ===== END ===== ===== MODEL INVENTORY ===== ===== REDPOINT OLLAMA MODEL INVENTORY LATEST ===== STAMP=2026-07-15_23:40:29 BACKEND_COUNT=16 REACHABLE_COUNT=13 UNIQUE_MODEL_COUNT=7 ===== MODEL PRESENCE ===== qwen2.5:1.5b -> n205,n22,n23,n254,n29,n30,n31 qwen2.5:3b -> n31 qwen3.5:0.8b -> n254 qwen3.5:4b -> n254,n31 qwen3.5:9b -> n254,n31 qwen3:1.7b -> n201,n203,n204,n206,n207,n209 qwen3:14b -> n209,n254 ===== NODE MODEL INVENTORY ===== n201 UP models=1 role=pve url=http://192.168.0.201:11434 qwen3:1.7b n203 UP models=1 role=pve url=http://192.168.0.203:11434 qwen3:1.7b n204 UP models=1 role=pve url=http://192.168.0.204:11434 qwen3:1.7b n205 UP models=1 role=pve url=http://192.168.0.205:11434 qwen2.5:1.5b n206 UP models=1 role=pve url=http://192.168.0.206:11434 qwen3:1.7b n207 UP models=1 role=pve url=http://192.168.0.207:11434 qwen3:1.7b n209 UP models=2 role=pve-ai url=http://127.0.0.1:11436 qwen3:1.7b, qwen3:14b n21 DOWN models=0 role=windows url=http://192.168.0.21:11434 error=URLError(TimeoutError('timed out')) n22 UP models=1 role=windows url=http://192.168.0.22:11434 qwen2.5:1.5b n23 UP models=1 role=windows url=http://192.168.0.23:11434 qwen2.5:1.5b n254 UP models=5 role=windows-m2 url=http://192.168.0.254:11434 qwen2.5:1.5b, qwen3.5:0.8b, qwen3.5:4b, qwen3.5:9b, qwen3:14b n26 DOWN models=0 role=windows-fast url=http://192.168.0.26:11434 error=URLError(OSError(113, 'No route to host')) n27 DOWN models=0 role=windows-fast url=http://192.168.0.27:11434 error=URLError(OSError(113, 'No route to host')) n29 UP models=1 role=windows-fast url=http://192.168.0.29:11434 qwen2.5:1.5b n30 UP models=1 role=windows url=http://192.168.0.30:11434 qwen2.5:1.5b n31 UP models=4 role=windows url=http://192.168.0.31:11434 qwen2.5:1.5b, qwen2.5:3b, qwen3.5:4b, qwen3.5:9b ===== ALIAS TARGET VALIDATION ===== rp-fast:latest BAD n254/qwen2.5:1.5b ok n26/qwen2.5:1.5b missing_or_unreachable n27/qwen2.5:1.5b missing_or_unreachable n29/qwen2.5:1.5b ok rp-fast-wide:latest BAD n254/qwen2.5:1.5b ok n26/qwen2.5:1.5b missing_or_unreachable n27/qwen2.5:1.5b missing_or_unreachable n29/qwen2.5:1.5b ok n23/qwen2.5:1.5b ok rp-all-qwen25:latest BAD n254/qwen2.5:1.5b ok n26/qwen2.5:1.5b missing_or_unreachable n27/qwen2.5:1.5b missing_or_unreachable n29/qwen2.5:1.5b ok n23/qwen2.5:1.5b ok n205/qwen2.5:1.5b ok n21/qwen2.5:1.5b missing_or_unreachable n22/qwen2.5:1.5b ok n30/qwen2.5:1.5b ok n31/qwen2.5:1.5b ok rp-254-fast:latest OK n254/qwen2.5:1.5b ok rp-14b:latest OK n254/qwen3:14b ok n209/qwen3:14b ok rp-qwen35-9b:latest BAD n254/qwen3.5:9b ok n26/qwen3.5:9b missing_or_unreachable n27/qwen3.5:9b missing_or_unreachable n31/qwen3.5:9b ok ===== MODEL PROFILES ===== fast_default: qwen2.5:1.5b qwen2.5:1.5b: n205,n22,n23,n254,n29,n30,n31 light_thinking: qwen3:1.7b qwen3:1.7b: n201,n203,n204,n206,n207,n209 mid_manual: qwen3.5:4b, qwen2.5:3b qwen3.5:4b: n254,n31 qwen2.5:3b: n31 heavy_manual: qwen3.5:9b, qwen3:14b qwen3.5:9b: n254,n31 qwen3:14b: n209,n254 ===== FUTURE NODE PULL PLAN ===== For any newly activated Windows worker, minimum seed model: ollama pull qwen2.5:1.5b For stronger Windows/GPU workers, optional: ollama pull qwen3.5:4b ollama pull qwen3.5:9b For large/manual nodes only: ollama pull qwen3:14b ===== FUTURE CANDIDATES ===== n20 state=offline_or_blocked url=http://192.168.0.20:11434 n24 state=offline_or_blocked url=http://192.168.0.24:11434 n25 state=credential_rejected url=http://192.168.0.25:11434 n28 state=offline_or_blocked url=http://192.168.0.28:11434 ALIAS_VALIDATION_BAD=4 NOTE=No models were pulled and no routing was changed. RESULT=MODEL_INVENTORY_OK ===== END =====