摘要:
A computing system includes a chassis, one or more backplanes coupled to the chassis. Computing devices are coupled to the one or more backplanes. The one or more backplanes include backplane openings that allow air to pass from one side of the backplane to the other side of the backplane. Air channels are formed by adjacent circuit board assemblies of the computing devices and the one or more backplanes. Channel capping elements at least partially close the air channels.
摘要:
Methods and apparatus for bandwidth metering in large-scale networks are disclosed. Metadata for a network transmission involving a virtualized resource at a host of a provider network, including endpoint address information and a traffic metric, is determined at a metering component. The metadata is aggregated at another metering component and provided to a traffic classification node. The traffic classification node generates a categorized usage record for the network transmission, based at least in part on network topology information associated with the provider network. The categorized usage record is used to determine a billing amount for the network transmission.
摘要:
Distributed database management systems may maintain collections of items spanning multiple partitions. Index structures may correspond to items on one partition or to items on multiple partitions. Item collections and indexes may be replicated. Changes to the data maintained by the distributed database management system may result in updates to multiple index structures. The changes may be compiled into an instruction set applicable to the index structures. In-memory buffers may contain the instructions prior to transmission to affected partitions. Replication logs may be combined with an acknowledgment mechanism for reliable transmission of the instructions to the affected partitions.
摘要:
Nodes of a database service may receive a read request to perform a read of a record stored by the database service and a transaction request to perform a transaction to the record. First and second indications of time may be associated with the read and transaction, respectively. A potential read anomaly (e.g., fuzzy read, read skew, etc.) may be detected based, at least in part, on a determination that the first indication of time is within a threshold value of the second indication of time. In response to detecting the potential read anomaly, the read may be performed after the transaction specified by the transaction request, regardless of whether the first indication of time is indicative of an earlier point in time than the second indication of time.
摘要:
Data is encrypted such that multiple keys are needed to decrypt the data. The keys are accessible to different entities so that no single entity has access to all the keys. At least one key is managed by a service provider. A customer computer system of the service provider may be configured with executable instructions directing the orchestration of communications between the various entities having access to the keys. As a result, security compromise in connection with a key does not, by itself, render the data decryptable.
摘要:
Techniques, including systems and methods, for capturing data sets include performing a client-side two-phase commit to ensure one or more data consistency conditions. A logical volume may represent a data set that is distributed among a plurality of physical storage devices. One or more client devices are instructed to block at least acknowledgment of write operations. When the one or more client devices have blocked at least acknowledgment of write operations, one or more servers in communication with the physical storage devices are instructed to capture corresponding portions of the data set. When the servers have been instructed to capture corresponding portions of the data set, the client devices are instructed to resume at least acknowledgment of write operations.
摘要:
A distributed lock manager (DLM) may be implemented in a distributed computing system that provides virtualized computing resources and/or virtual computing services to clients. Locks may be created and managed by the DLM in response to requests from clients to do so. The components of the DLM may communicate with each other over a network other than one over which client application components communicate with each other or access the shared resources protected by the locks. For example, DLM components may communicate over a control plane network of a cloud computing environment and application components may communicate over a data plane network of the cloud computing environment. The DLM may expose an API to clients, allowing them to make local calls to a DLM component on the same node to perform various lock management operations. The meanings of the lock values may be established by their use in client applications.