An official website of the United States government
Noria, first presented at OSDI 2018, is a new web application backend that delivers the same fast reads as an in-memory cache in front of the database, but without the application having to manage the cache. Even better, Noria still accepts SQL queries and allows changes to the queries without extra effort, just like a database. Noria performs well: it serves up to 14M requests per second on a single server, and supports a 5x higher load than carefully hand-tuned queries issued to MySQL. Writing web applications that tolerate high load is difficult. The reason is that the backend storage system that the application relies on—typically a relational database, like MySQL— can easily become a serious bottleneck with many clients. Each page view typically involves 10 or more database queries, which each take up CPU time on the database servers to evaluate. To avoid such slow database interactions and to reduce load on the database, applications often introduce caches (like memcached or Redis) that store already-computed query results for fast common case access. These caches, however, impose significant application complexity, because the application must query, invalidate, and maintain them. Surely there has to be a better way.
more »
« less
Noria: A New Take on Fast Web Application Backends
Noria, first presented at OSDI ’18, is a new web application backend that delivers the same fast reads as an in-memory cache in front of the database, but without the application having to manage the cache. Even better, Noria still accepts SQL queries and allows changes to the queries without extra effort, just like a database. Noria performs well: it serves up to 14M requests per second on a single server, and supports a 5x higher load than carefully hand-tuned queries issued to MySQL.
more »
« less
- Award ID(s):
- 1704376
- PAR ID:
- 10129101
- Date Published:
- Journal Name:
- Login
- Volume:
- 44
- Issue:
- 1
- ISSN:
- 1044-6397
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Side-channel attacks leverage implementation of algorithms to bypass security and leak restricted data. A timing attack observes differences in runtime in response to varying inputs to learn restricted information. Most prior work has focused on applying timing attacks to cryptoanalysis algorithms; other approaches sought to learn about database content by measuring the time of an operation (e.g., index update or query caching). Our goal is to evaluate the practical risks of leveraging a non-privileged user account to learn about data hidden from the user account by access control. As with other side-channel attacks, this attack exploits the inherent nature of how queries are executed in a database system. Internally, the database engine processes the entire database table, even if the user only has access to some of the rows. We present a preliminary investigation of what a regular user can learn about “hidden” data by observing the execution time of their queries over an indexed column in a table. We perform our experiments in a cache-control environment (i.e., clearing database cache between runs) to measure an upper bound for data leakage and privacy risks. Our experiments show that, in a real system, it is difficult to reliably learn about restricted data due to natural operating system (OS) runtime fluctuations and OS-level caching. However, when the access control mechanism itself is relatively costly, a user can not only learn about hidden data but they may closely approximate the number of rows hidden by the access control mechanism.more » « less
-
Cormode, Graham; Shekelyan, Michael (Ed.)Data analytics skills have become an indispensable part of any education that seeks to prepare its students for the modern workforce. Essential in this skill set is the ability to work with structured relational data. Relational queries are based on logic and may be declarative in nature, posing new challenges to novices and students. Manual teaching resources being limited and enrollment growing rapidly, automated tools that help students debug queries and explain errors are potential game-changers in database education. We present a suite of tools built on the foundations of database theory that has been used by over 1600 students in database classes at Duke University, showcasing a high-impact application of database theory in database education.more » « less
-
RDBMSes only support one clustered index per database table that can speed up query processing. Database applications, that continually ingest large amounts of data, perceive slow query response times to long downtimes, as the clustered index ordering must be strictly maintained. In this paper, we show that application slowdown or downtime, however, can often be avoided if database systems expose the physical location of attributes that are completely or approximately clustered. Towards this, we propose PLI, a physical location index, constructed by determining the physical ordering of an attribute and creating approximately sorted buckets that map physical ordering with attribute values in a live database. To use a PLI incoming SQL queries are simply rewritten with physical ordering information for that particular database. Experiments show queries with the PLI index significantly outperform queries using native unclustered (secondary) indexes, while the index itself requires a much lower maintenance overheads when compared to native clustered indexes.more » « less
-
A multiverse database transparently presents each application user with a flexible, dynamic, and independent view of shared data. This transformed view of the entire database contains only information allowed by a centralized and easily-auditable privacy policy. By enforcing the privacy policy once, in the database, multiverse databases reduce programmer burden and eliminate many frontend bugs that expose sensitive data. Multiverse databases' per-user transformations risk expensive queries if applied dynamically on reads, or impractical storage requirements if the database proactively materializes policy-compliant views. We propose an efficient design based on a joint dataflow across "universes" that combines global, shared computation and cached state with individual, per-user processing and state. This design, which supports arbitrary SQL queries and complex policies, imposes no performance overhead on read queries. Our early prototype supports thousands of parallel universes on a single server.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- The DOI is not currently available.
