CLAUDE LABJP
OPUS48 — The default model on Bedrock, Vertex, and AWS is now Claude Opus 4.8. Opus 4.8 and Haiku 4.5 are also in the Messages API, with prompt caching and extended thinkingSTREAM — A Claude Code stability and quality update landed: subagent text over stream-json, stronger permission and hook handling, and better background-agent reportingFASTEND — Fast mode for Claude Opus 4.7 is removed on July 24. After that, speed: 'fast' returns an error, so migrate to fast mode on Opus 4.8TEACH — Claude for Teachers is here, giving verified US K-12 educators free premium access, plus curriculum connections aligned to standards in all 50 states and new education connectorsFIX — The update fixes issues across Chrome, Windows, Bedrock, Vertex, hooks, and session recovery, and speeds up terminal renderingIPO — Bankers are reportedly lining up investor meetings for Anthropic ahead of a possible public listing as soon as OctoberOPUS48 — The default model on Bedrock, Vertex, and AWS is now Claude Opus 4.8. Opus 4.8 and Haiku 4.5 are also in the Messages API, with prompt caching and extended thinkingSTREAM — A Claude Code stability and quality update landed: subagent text over stream-json, stronger permission and hook handling, and better background-agent reportingFASTEND — Fast mode for Claude Opus 4.7 is removed on July 24. After that, speed: 'fast' returns an error, so migrate to fast mode on Opus 4.8TEACH — Claude for Teachers is here, giving verified US K-12 educators free premium access, plus curriculum connections aligned to standards in all 50 states and new education connectorsFIX — The update fixes issues across Chrome, Windows, Bedrock, Vertex, hooks, and session recovery, and speeds up terminal renderingIPO — Bankers are reportedly lining up investor meetings for Anthropic ahead of a possible public listing as soon as October
TAG

retry storm

1 articles
Back to all tags
Related:
Claude API1tail latency1production1scheduled workload1indie developer1
API & SDK/2026-05-27Advanced

Tail Latency in Scheduled Claude API Workloads: A Three-Layer Guardrail Against Retry Storms

After running six sites in parallel through scheduled Claude API tasks for several months, 14 days of logs revealed three distinct p95/p99 patterns and a retry storm I had been creating from my own client. This is the guardrail design I landed on — jitter, budget, circuit breaker — with the before/after numbers.