Amazon customers who have had the failure to meet. All clients in the earlier this week affected data center may be eligible for compensation. Customers who have data loss and have had to get extra compensation.
Amazon was the beginning of the last week, embarrassed by a large-scale failure in its data centre in Dublin creating a 110kV 10MW transformer without power. Although initially it was thought that this was caused by lightning, the energy company now going out of a different cause. About the possible cause are no details released, other than that the power company is still investigating the cause.
The hosting company has, since last Friday the failure and the aftermath under control and subsequently this weekend detailed information is released about the progress of the fault. Also has Amazon a compensation scheme drawn up which details are published. All customers in the affected Availability Zone a EBS volume or RDS database in use, ten days ‘ use of their EBS volume to EC2 instance and RDS database instance in the affected Availability Zone offset. It doesn’t matter whether a customer was actually affected by the consequences of the failure or not.
In addition, the Amazon, a more extensive compensation scheme drawn up for customers who have been affected by what the company EBS-softwarebug calls. This bug is according to Amazon release of the blackout, and came the day before the power failure to light. By the bug saw the Amazon software certain opslagblokken as unused in the list were set to be deleted. A second run of this process, they discovered the error, however, and marked this for further research. The Amazon employee who subsequently these highlighted items had to investigate, however, seems not to have looked at which data blocks still were removed.
Clients who receive this error are affected will receive 30 days of free use of their EBS usage in all Availability Zones in the so-called EU West Region. The free use includes both snapshots and requests as volume storage and I/O. in Addition, this group of customers in that period free of charge Premium Support via the AWS Support Center.
Amazon says several measures have been taken to its infrastructure to improve or there about to consult with manufacturers. This has included the uninstall process honed to the EBS-bug crush, and is an additional way of snapshot-recovery to be rolled out. In the current setup, a volume can only be restored if the data on the EBS server is first moved. This allows for significant delay because there are sometimes large amounts of data need to be copied. This problem wants Amazon to work around the option to offer a volume immediately after a power failure can be restored without data having to move.