Amazon.com Inc. has published an apology for troubles at its data center which took place on Friday. In the apology it admits that one of its data-center faced outage which consequently led to several widely used websites to fall down. Some of the key websites affected by this mishap were Foursquare and Reddit. The apology also went on announcing that as compensation Amazon Inc. has granted free credit of 10 day to all its Web services customers which were affected.
Image via CrunchBase The company has not yet declared the degree of loss, or the amount of credits this incident has cost. Amazon Web Services is not exactly the best-money making business of Amazon Inc., in fact it only counts few percent of the total annual revenue of Amazon. However, having said that, recently Amazon Inc. is working hard on its Web Services and aims to have better results from the business in the future. The service has been renting out computers on hourly basis.
The incident occurred at data center which was located next to Dulles Airport, outside Washington. This was first of its magnitude and a big mishap for the service which affect its repute adversely. Amazon is striving tirelessly in order to restore the computers in the stable position they were before the incident, these efforts are underway since eight days now. According to the result of company’s internal investigation it reported on Friday that the hindrance in the system was generated by a human error which led to the outage. Right after which, according to the protocol, an automated process of error-recovery was started which unfortunately went uncontrollable; hence this resulted in several computers at the server to becoming "stuck" while still in the recovery mode.
The servers of Amazon are designed to work best without any human intervention as they work on their own, this is achieved by different computers allowed work on a different "availability zone," and one substitutes the other if the first one stops functioning. Amazon informed according to its investigation, that those customers who had services computing tasks to run over more than one zones remained unaffected to some extent. However, it also agreed that the error also resulted in a degree of difficulty for these websites to switch zones on the fly and it also promised that it is trying its best to make amendments in the system to avoid any further such error from now onwards.
The credit awarded as a condensation has been given to all the customers using that particular zone which was halted, regardless of whether or not that website was directly affected. Amazon is not disclosing the exact number or list of customers which were actually affected by this mishap at their data center.
No comments:
Post a Comment