Introducing the Server Side Cloud Swift SDK

Cloudflare has closed a network-performance incident affecting Madrid after reporting a fix on September 30. Its incident timeline records monitoring from 14:06 UTC that day and a resolved update at 05:42 UTC on October 2. The provider classified the incident as minor; its notice does not identify a root cause or quantify customer impact.

Cloudflare marked its Workers build-delay incident resolved at 21:27 UTC on October 1. The company had opened an investigation at 15:33 UTC on September 30, warning of potential effects on multiple customers, then moved to monitoring after implementing a fix at 17:42 UTC on September 30.

Cloudflare reports that an incident involving increased Durable Objects and R2 errors in Amsterdam was resolved on October 1 at 21:18 UTC. Its status history records an investigation beginning at 20:59 UTC and a fix entering monitoring at 21:06 UTC. Those updates do not identify a root cause or quantify affected customers.

In a Google Cloud customer account, PayPal describes migrating analytics workloads from on-premises Hadoop systems to Managed Service for Apache Spark. The company says the move consolidated fragmented workflows, shortened cluster provisioning and improved integration with Cloud Storage and BigQuery. PayPal reports a 25% improvement in core workload processing times, alongside lower operational overhead; those results reflect its own deployment.

Google says the latest Cloud Storage SDKs now enable end-to-end upload checksumming by default, reducing extra work previously required from application developers. The SDKs calculate and send checksums when applications do not supply them, and support verification during downloads. Range reads through the gRPC API also receive integrity checks. Google recommends updating clients to use these protections against accidental data corruption.

Google Cloud is positioning Managed Lustre’s Dynamic Tier as shared storage for AI training and high-performance computing, combining a fast cache with a larger capacity pool. The update also targets developer workloads such as repository operations and library loading, aiming to reduce separate storage environments. Published performance comparisons are Google’s tests, with test configurations supplied for evaluation.