The extent of the impact of the DDoS assault could not be immediately determined. However, Allen Goldberg, spokesman for UltraDNS, said the attack started about 4:45 p.m. Pacific time Wednesday, causing some delays to customers.
UltraDNS had its systems running normally within an hour, and the attack only affected Northern California Internet users, Goldberg said.
"No one was ever out. There was no downtime," Goldberg told InformationWeek Thursday. "Queries may have taken some time to resolve and some may not have been completed, but there was never an outage."
Details of the attack were not available, and Goldberg declined to discuss specific customers. However, postings on the Twitter thread of Jeff Barr, a strategist for Amazon Web Services, indicated that AWS and Amazon were affected. Amazon could not be reached for comment in time for this writing.
The AWS Service Health Dashboard indicated that AWS' storage and computing cloud services, S3 and EC2, respectively, were affected. At 5:44 p.m. Pacific AWS was investigating reports of "DNS resolution errors," and by 6:02 p.m. confirmed that "some customers in the West Coast are experiencing issues with resolving DNS." By 6:39 p.m., the system had fully recovered.
DDoS attacks occur regularly on the Web and are usually brought under control by service providers before the assaults cause serious damage. AWS has been a target of DDoS attacks before.
A suspected denial-of-service attack on AWS shut down a code hosting service for nearly 24 hours in early October.