Cloudflare degraded D1, Durable Objects, and R2 in Hong Kong and Singapore for four and a half hours

Sources: Cloudflare incident page, Cloudflare incidents feed
Summary: Cloudflare reports errors and elevated latency for D1, Durable Objects, and R2 in the Hong Kong and Singapore colos from 03:53 to 08:28 UTC on 2026-08-24, with a fix implemented at 06:42 UTC. Cloudflare rates the impact minor and has published no root cause. The same incidents feed carries two prior degradations in the same product family within four days: Hong Kong Durable Objects and R2 from 14:59 to 15:19 UTC on 2026-08-21, and Dubai R2 on 2026-08-22. Cloudflare rates all three minor and has published no root cause for any of them, so no common cause is established.
Why it matters: Workers applications reading or writing D1 databases or Durable Objects pinned to the Hong Kong and Singapore colos saw errors across a four and a half hour window, and the storage products behind them have now degraded three times in four days without a published cause.
Follow-up: Track whether Cloudflare publishes a root cause covering the 2026-08-21, 2026-08-22, and 2026-08-24 storage degradations.