perf(tracing): skip span-start upsert by default (end-only ingest) #1647

Job	Run time
find-tutorials	4s
test-00_sync/040_pydantic_ai	2m 12s
test-00_sync/050_openai_agents_local_sandbox	2m 23s
test-10_async/00_base/030_tracing	2m 12s
test-00_sync/030_langgraph	3m 12s
test-10_async/00_base/080_batch_events	3m 20s
test-00_sync/010_multiturn	2m 56s
test-00_sync/000_hello_acp	2m 8s
test-10_async/00_base/120_openai_agents_local_sandbox	4m 25s
test-10_async/00_base/010_multiturn	2m 29s
test-00_sync/020_streaming	3m 42s
test-10_async/10_temporal/050_agent_chat_guardrails	1m 57s
test-10_async/10_temporal/070_open_ai_agents_sdk_tools	2m 9s
test-10_async/00_base/020_streaming	2m 11s
test-10_async/10_temporal/000_hello_acp	2m 15s
test-10_async/00_base/000_hello_acp	2m 9s
test-10_async/10_temporal/010_agent_chat	3m 4s
test-10_async/10_temporal/020_state_machine	2m 45s
test-10_async/00_base/040_other_sdks	3m 59s
test-10_async/10_temporal/120_openai_agents_local_sandbox	2m 15s
test-10_async/10_temporal/030_custom_activities	2m 7s
test-10_async/00_base/100_langgraph	2m 25s
test-10_async/10_temporal/110_pydantic_ai	2m 6s
test-10_async/10_temporal/060_open_ai_agents_sdk_hello_world	2m 10s
test-10_async/10_temporal/080_open_ai_agents_sdk_human_in_the_loop	2m 19s
test-10_async/00_base/110_pydantic_ai	2m 27s
test-10_async/10_temporal/090_claude_agents_sdk_mvp	2m 16s
test-10_async/10_temporal/130_langgraph	6m 28s
Test Summary	7s
	1h 14m 12s

Provide feedback