Archived edition · Monday 31 Aug 2026

Drama index 72/10010 degradedMood: Gusty

Discord turns gusty for some servers

Discord has left some servers out of reach, like a remote slipping beneath the sofa at precisely the wrong moment.

Render wobbles in Singapore as OpenAI adds latency to the agenda.Clerk reports elevated API latency; requests are now fashionably late.Cloudflare Access has availability issues; even the bypass needs a bypass.Hetzner load balancers are under maintenance; every location is redecorating.Hetzner load balancers are under maintenance; every location is redecorating.Zoom recovered after 85h 40m; EU billing may now resume its paperwork.

Should you deploy right now?

Ship itAbsolutely not

Use caution. Cloudflare is degraded.

Deployment inputs: 1 degraded · no unconfirmed warnings.

Top HN thread: I think the military commissary's freezers were hacked

DiscordGustySome servers not available for some usersDetails →

SendGridHazyInconsistencies with Suppressions - Block ListsDetails →

TwilioSlowInconsistencies with Suppressions - Block ListsDetails →

CloudflareSlowHTTP/3 issue affecting R2 custom domainsDetails →

SourcegraphDrizzleSourcegraph Cloud is experiencing elevated errorDetails →

KrakenDrizzleInjective (INJ) Funding DelaysDetails →

Cisco WebexDelayedWebex Calling: Some Devices are not visible in Collaboration Control HubDetails →

CoinbaseSlowPaused Sends/Receives - Core DAO NetworkDetails →

SnowflakeHazyService incidentDetails →

falSlowHistorical Request lookup and Assets generated between Aug 1 - 26 may be degradedDetails →

All clear · 119 services

Regional

AWS · 10 regions

Global us-east-1 ap-northeast-1 ap-south-1 ap-southeast-1 eu-central-1 eu-west-1 me-central-1 sa-east-1 us-west-2

Quiet · 32 community services with no reports

Forecast

Breaking: Discord reports some servers unavailable for some users, a fresh loose tile in an already busy hallway. Since the prior edition, Discord has moved from clear to degraded. Cloudflare continues to report an HTTP/3 issue affecting R2 custom domains, while Snowflake, SendGrid, and other services retain active incident notes. Several source-quality checks also remain unavailable, so some labels are less useful than the devices they describe. Market mood reads Greed, which has not volunteered to fix anything. Keep an eye on the dashboard.

Latest readings

22:45 UTCSentry recovered: was degraded for 2h 13m
22:05 UTCMailgun recovered: was degraded for 7h 17m
22:00 UTCNetlify recovered: was degraded for 1h 19m
20:40 UTCAnthropic API recovered: was degraded for 1h 18m
20:40 UTCClaude Code recovered: was degraded for 1h 18m
20:40 UTCNetlify degraded: Increased Latency in IAD Region on Global Network
20:35 UTCKlarna recovered: was degraded for 1h 15m
20:35 UTCSentry degraded: Ingestion backlog

Warnings in effect

Use a backup channel for Discord; the sofa has claimed part of the remote.

Advice for the day

Check whether your Discord server is affected before changing plans.
Keep deployment notes somewhere reachable without Discord.
Treat unavailable source checks as a label with the batteries removed.