# HELP ailix_active_streams Currently running durable LLM streaming tasks. # TYPE ailix_active_streams gauge ailix_active_streams 0.0 # HELP ailix_stream_first_token_seconds Seconds from worker start to first provider token. # TYPE ailix_stream_first_token_seconds histogram ailix_stream_first_token_seconds_bucket{le="0.005"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.01"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.025"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.05"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.075"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.1"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.25"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.5"} 0.0 ailix_stream_first_token_seconds_bucket{le="0.75"} 0.0 ailix_stream_first_token_seconds_bucket{le="1.0"} 0.0 ailix_stream_first_token_seconds_bucket{le="2.5"} 0.0 ailix_stream_first_token_seconds_bucket{le="5.0"} 0.0 ailix_stream_first_token_seconds_bucket{le="7.5"} 0.0 ailix_stream_first_token_seconds_bucket{le="10.0"} 0.0 ailix_stream_first_token_seconds_bucket{le="+Inf"} 0.0 ailix_stream_first_token_seconds_count 0.0 ailix_stream_first_token_seconds_sum 0.0 # HELP ailix_stream_first_token_seconds_created Seconds from worker start to first provider token. # TYPE ailix_stream_first_token_seconds_created gauge ailix_stream_first_token_seconds_created 1.7862703105010443e+09 # HELP ailix_stream_total_seconds Total seconds spent by a durable LLM streaming task. # TYPE ailix_stream_total_seconds histogram # HELP ailix_redis_xadd_seconds Seconds spent appending a chat streaming event to Redis Streams. # TYPE ailix_redis_xadd_seconds histogram ailix_redis_xadd_seconds_bucket{le="0.005"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.01"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.025"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.05"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.075"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.1"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.25"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.5"} 0.0 ailix_redis_xadd_seconds_bucket{le="0.75"} 0.0 ailix_redis_xadd_seconds_bucket{le="1.0"} 0.0 ailix_redis_xadd_seconds_bucket{le="2.5"} 0.0 ailix_redis_xadd_seconds_bucket{le="5.0"} 0.0 ailix_redis_xadd_seconds_bucket{le="7.5"} 0.0 ailix_redis_xadd_seconds_bucket{le="10.0"} 0.0 ailix_redis_xadd_seconds_bucket{le="+Inf"} 0.0 ailix_redis_xadd_seconds_count 0.0 ailix_redis_xadd_seconds_sum 0.0 # HELP ailix_redis_xadd_seconds_created Seconds spent appending a chat streaming event to Redis Streams. # TYPE ailix_redis_xadd_seconds_created gauge ailix_redis_xadd_seconds_created 1.7862703105011818e+09 # HELP ailix_redis_xread_seconds Seconds spent waiting for a chat streaming Redis XREAD call. # TYPE ailix_redis_xread_seconds histogram ailix_redis_xread_seconds_bucket{le="0.005"} 0.0 ailix_redis_xread_seconds_bucket{le="0.01"} 0.0 ailix_redis_xread_seconds_bucket{le="0.025"} 0.0 ailix_redis_xread_seconds_bucket{le="0.05"} 0.0 ailix_redis_xread_seconds_bucket{le="0.075"} 0.0 ailix_redis_xread_seconds_bucket{le="0.1"} 0.0 ailix_redis_xread_seconds_bucket{le="0.25"} 0.0 ailix_redis_xread_seconds_bucket{le="0.5"} 0.0 ailix_redis_xread_seconds_bucket{le="0.75"} 0.0 ailix_redis_xread_seconds_bucket{le="1.0"} 0.0 ailix_redis_xread_seconds_bucket{le="2.5"} 0.0 ailix_redis_xread_seconds_bucket{le="5.0"} 0.0 ailix_redis_xread_seconds_bucket{le="7.5"} 0.0 ailix_redis_xread_seconds_bucket{le="10.0"} 0.0 ailix_redis_xread_seconds_bucket{le="+Inf"} 0.0 ailix_redis_xread_seconds_count 0.0 ailix_redis_xread_seconds_sum 0.0 # HELP ailix_redis_xread_seconds_created Seconds spent waiting for a chat streaming Redis XREAD call. # TYPE ailix_redis_xread_seconds_created gauge ailix_redis_xread_seconds_created 1.7862703105012655e+09 # HELP ailix_redis_xadd_to_xread_seconds Seconds between Redis Stream event ID time and API XREAD delivery. # TYPE ailix_redis_xadd_to_xread_seconds histogram ailix_redis_xadd_to_xread_seconds_bucket{le="0.005"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.01"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.025"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.05"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.075"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.1"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.25"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.5"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="0.75"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="1.0"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="2.5"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="5.0"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="7.5"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="10.0"} 0.0 ailix_redis_xadd_to_xread_seconds_bucket{le="+Inf"} 0.0 ailix_redis_xadd_to_xread_seconds_count 0.0 ailix_redis_xadd_to_xread_seconds_sum 0.0 # HELP ailix_redis_xadd_to_xread_seconds_created Seconds between Redis Stream event ID time and API XREAD delivery. # TYPE ailix_redis_xadd_to_xread_seconds_created gauge ailix_redis_xadd_to_xread_seconds_created 1.7862703105013149e+09 # HELP ailix_zombie_messages_reaped_total Stale streaming message rows marked failed by the RS.6 reaper. # TYPE ailix_zombie_messages_reaped_total counter ailix_zombie_messages_reaped_total 0.0 # HELP ailix_zombie_messages_reaped_created Stale streaming message rows marked failed by the RS.6 reaper. # TYPE ailix_zombie_messages_reaped_created gauge ailix_zombie_messages_reaped_created 1.786270310501396e+09 # HELP ailix_stream_errors_total Durable LLM streaming failures by phase and provider. # TYPE ailix_stream_errors_total counter # HELP ailix_litellm_gateway_ready Whether the LiteLLM staging gateway is configured and enabled. # TYPE ailix_litellm_gateway_ready gauge ailix_litellm_gateway_ready 1.0 # HELP ailix_litellm_proxy_requests_total LiteLLM staging proxy requests by result. # TYPE ailix_litellm_proxy_requests_total counter # HELP ailix_provider_attempts_total Provider runtime attempts by effective provider, status, and fallback kind. # TYPE ailix_provider_attempts_total counter # HELP ailix_provider_attempt_first_token_seconds Seconds from provider attempt start to first token. # TYPE ailix_provider_attempt_first_token_seconds histogram # HELP ailix_provider_attempt_total_seconds Total seconds spent by a provider runtime attempt. # TYPE ailix_provider_attempt_total_seconds histogram # HELP ailix_provider_gateway_errors_total Provider/gateway failures by phase and normalized failure class. # TYPE ailix_provider_gateway_errors_total counter # HELP ailix_provider_fallbacks_total Runtime fallback attempts by fallback kind and provider. # TYPE ailix_provider_fallbacks_total counter # HELP ailix_trace_correlation_failures_total Best-effort runtime trace correlation failures by sink. # TYPE ailix_trace_correlation_failures_total counter # HELP ailix_stream_reconnect_observed_total SSE reconnects observed with a Last-Event-ID cursor. # TYPE ailix_stream_reconnect_observed_total counter ailix_stream_reconnect_observed_total 0.0 # HELP ailix_stream_reconnect_observed_created SSE reconnects observed with a Last-Event-ID cursor. # TYPE ailix_stream_reconnect_observed_created gauge ailix_stream_reconnect_observed_created 1.786270310501596e+09 # HELP ailix_cache_hit_rate Prompt-cache hit rate (0..1) per route/provider over the last 7 days. # TYPE ailix_cache_hit_rate gauge # HELP ailix_cache_canary_degraded_total Times a route's week-over-week cache hit rate collapsed past the canary threshold. # TYPE ailix_cache_canary_degraded_total counter