Amazon’s Simple Queue Service (SQS) provides durable messaging guarantees and is an excellent backbone for messaging services. However, SQS does not support “fan-out” of messages so that multiple consuming services can each receive a copy of a message. This means that true publish-subscribe messaging requires some additional work. This post describes some architectural choices that provide durable publish-subscribe messaging using SQS by tracking messaging subscribers using a database, and matching published messages to interested subscribers. ...

Title and Author of Paper The Anatomy of a Large-Scale Hypertextual Web Search Engine. Sergey Brin and Lawrence Page. Summary This paper describes the underpinnings of the Google search engine. The paper presents the initial Google prototype and describes the challenges in scaling search engine technology to handle large datasets. At the time of writing, the main goal of Google is to improve the quality of web searches by taking advantage of the existing link data embedded in web pages to calculate the quality of a page. ...

Title and Author of Paper Consistency Analysis in Bloom: a CALM and Collected Approach. Alvaro et al. Summary Distributed programming is difficult for even experienced developers to get correct. Understanding the tradeoff between consistency, availability, and latency, while guaranteeing data correctness, provides a wealth of problems for the application developer. This paper presents a language and method for programmatically verifying distributed consistency. CALM - Consistency and Logical Monotonicity There is a connection between distributed consistency algorithms and logical monotonicity, that is, our programs must be correct even in the face of the delay and re-ordering of messages and data across different nodes in a system. By designing programs with order independence, we can ensure that the logic of the system is maintained within a distributed environment. ...

When designing Workiva’s durable messaging system we took a hard look at using Amazon’s Kinesis as the message storage and delivery mechanism. At first glance, Kinesis has a feature set that looks like it can solve any problem: it can store terabytes of data, it can replay old messages, and it can support multiple message consumers. But if you dig a little deeper you will find that Kinesis is well suited for a very particular use case, and if your application doesn’t fit this use case, Kinesis may be a lot more trouble than it’s worth. ...

Architecture is the bridge between (often abstract) business goals and the final (concrete) resulting system. – Software Architecture in Practice A software architect should act as a bridge between business stakeholders and technical stakeholders. To be this bridge requires understanding the business problem being solved, and being able to distill that problem into a technical solution that a software team can implement. In essence, the architect acts as a technical business analyst that helps to define the needs of an organization and recommend solutions that deliver value to stakeholders. It just so happens that an architect’s stakeholders include project managers, software developers, management, and other software architects. This article describes how a software architect can apply methods from business analysis to make their job easier and to satisfy these different stakeholders. ...

Title and Author of Paper The CQL continuous query language: semantic foundations and query execution. Arasu et al. Summary CQL is a derivation of the SQL query language developed for running continuous queries over streams of data. The goals of the system are to provide a precise set of language semantics for running such continuous stream workloads. The paper starts by defining precise abstract semantics for continuous queries that cover two data types — streams and relations — and three classes of operators: ones that produce a relation from a stream, one that produces a relation from other relations, and one that produces a stream from a relation. These semantics are defined independent of the underlying implementation. The second portion of this paper defines how CQL instantiates these abstract semantics using existing SQL specifications and some new CQL additions. ...

Title and Author of Paper BlinkDB: Queries with Bounded Errors and Bounded Response Times on Very Large Data. Agarwal et al. Summary BlinkDB is a massively parallel database that provides approximate results for queries over large data sets. BlinkDB’s distinguishing feature is providing the opportunity for users to trade response time for query accuracy — partial results are returned with annotated error bars describing their accuracy at the current point in time. ...

Title and Author of Paper Informix under CONTROL: Online Query Processing. J. M. Hellerstein et al. Summary The CONTROL project attempts to improve the interaction between users and computers during data analysis. Traditional data analysis systems are a black box where a user enters a query, and waits for some amount of time before receiving a result. The CONTROL project aims to make this process interactive by continuously providing approximate results that are improved over time. Implementing such a system requires rethinking some fundamental tenants of database systems. First, with interactive systems queries may never complete, but instead they may be halted when results are “good enough”. Second, interactive systems must be able to provide approximate results quickly while maximizing the rate at which an accurate answer is found. This paper explores the changes in database technology needed to support interactive use cases. ...

One of the realities of running a business is that you will — at some point — need to make money. This means you must sell something for more than it cost to make it. Profit = Total Revenue - Total Expenses So, to sell a service or feature (like a third-party API portal) at a profit, we need to understand how much it costs to make it, so we know how much to sell it for. In business parlance, the quantity we are looking for is the cost of goods sold — the accumulated costs used to create a good, including direct labor, materials, and any overhead. ...

Title and Author of Paper An Array-Based Algorithm for Simultaneous Multidimensional Aggregates. Y. Zhao et al. Summary One of the core functions of an OLAP system is computing aggregations and group-by operations. This functionality has been characterized by the “Cube” operator, which computes group-by aggregations over all possible subsets of a specified dimension. As an example of the Cube operator, consider a model with the dimensions product, store, date, and the measured value sales. To compute the Cube for this data set requires computing sales for all subsets of the dimensions: sales by product, store, and date; sales by product and store; sales by product; etc. As a user, I want the system to prepare these results for me in response to ad-hoc queries or as part of a ETL job that prepares the data for analysis. Because there is a lot of data involved, the challenge of implementing the Cube operator is in computing these aggregations as efficiently as possible. ...