Tuning the buffer size appropriately is critical to the performance of a cloud database, since memory is usually the resource bottleneck. For large-scale databases supporting heterogeneous applications, configuring the individual buffer sizes for a significant number of database instances presents a scalability challenge. Manual optimization is neither efficient nor effective, and even not feasible for large cloud clusters, especially when the workload may dynamically change on each instance. The difficulty lies in the fact that each database instance requires a different buffer size that is highly individualized, subject to the constraint of the total buffer memory space. It is imperative to resort to algorithms that automatically orchestrate the buffer pool tuning for the entire database instances. To this end, we design iBTune that has been deployed for more than 10, 000 OLTP cloud database instances in our production system. Specifically, it leverages the information from similar workloads to find out the tolerable miss ratio of each instance. Then, it utilizes the relationship between miss ratios and allocated memory sizes to individually optimize the target buffer pool sizes. To provide a guaranteed level of service level agreement (SLA), we design a pairwise deep neural network that uses features from measurements on pairs of instances to predict the upper bounds of the request response times. A target buffer pool size can be adjusted only when the predicted response time upper bound is in a safe limit. The successful deployment on a production environment, which safely reduces the memory footprint by more than 17% compared to the original system that relies on manual configurations, demonstrates the effectiveness of our solution.

Most co-authoring tools support basic annotations, such as edits and comments that are anchored at specific locations in the document. However, they do not support meta-commentary about a document (such as an author's summary of modifications) which gets separated from the document, often in the body of email messages. This causes unnecessary overhead in the write-review-edit workflow inherent in co-authoring. We present document-embedded structured annotations called "bundles" that incorporate the meta-commentary into a unified annotation model that meets a set of annotation requirements we identified through a small field investigation. A usability study with 20 subjects evaluated the annotation reviewing stage of co-authoring and showed that annotation bundles in our high-fidelity prototype reduced reviewing time and increased accuracy, compared to a system that only supports edits and comments.