Amazon said in an update early Friday that meaningful progress has been made and is expected to continue over the next few hours.

    Reddit, a social news website, said on its homepage that "we are slowly getting our capacity（能力,容量） back," and users may be able to log in if lucky. Some sites appeared to have recovered, including location-based social networking site Foursquare and HootSuite, an online brand management and social websites monitoring service.

    Amazon has not released the number of companies which have been disrupted by the outage and when the web-services will fully recover.

    The outage highlights risks of relying on public cloud service. Security and reliability have been experts' concerns on public could web-hosting service. Many Internet startups have been relying on the service to bring their costs down.

    Elastic Compute Cloud (EC2) is a central part of Amazon's cloud computing platform Amazon Web Services. It allows users to rent virtual computers on which to run their own computer applications. The European Space Agency, PBS and Ericsson are all among its cloud service customers.

    According to Amazon's Service Health Dashboard, errors began at 1:41 a.m. PDT Thursday in its datacenter in northern Virginia, saying "instance connectivity, latency and error rates."(本文由在线英语听力室整理编辑)