Dell R720 ExpressFlash 350GB Leads SSDs in Overall Average Latency

MarkLogic Corporation, provider of ACID-compliant enterprise NoSQL (not only SQL) database solutions, and StorageReview.com, an independent storage authority, produced the first public results of StorageReview.com’s MarkLogic NoSQL Database Storage Benchmark.

                                 StorageReview.com MarkLogic NoSQL
                                       Database Storage Benchmark
                                     (Overall Average Latency in ms)

This test illustrates the performance of flash storage, hard drive arrays and PCIe application accelerators running on MarkLogic. The benchmark provides customers with access to the latest storage performance data to guide their selection of new storage solutions for their NoSQL deployments.

"Previous benchmarking efforts just simulated database workloads, but the MarkLogic benchmark provides us with the first opportunity to test an actual NoSQL instance working against the best host-side flash solutions on the market today," said Brian Beeler, editor in chief of StorageReview.com. "Our focus in this test is to look at overall latency from each examined storage solution across several performance factors."

The MarkLogic NoSQL database has the flexibility and scalability to handle today’s data challenges that SQL-based databases were not designed to handle. It has enterprise capabilities like search, ACID transactions, failover, replication, and security to run mission-critical applications. Any environment that faces the big data challenges of data volume, velocity, variety, and complexity can be enhanced with MarkLogic.

"This test enables our customers and prospects to evaluate the stratification in performance of various storage products being positioned as solutions for NoSQL environments," said Denis Sheahan, lead performance engineer at MarkLogic. "As a result, our customers will be able to easily select a storage solution qualified by StorageReview.com that meets their capacity, budgetary and/or performance needs."

The benchmark is used to evaluate hardware configurations as well as upcoming MarkLogic software releases.

The testing workload is divided into two distinct parts:

1. Ingestion phase where a large data set is inserted with indexes into the MarkLogic database.
2. Query phase where searches, views updates, and deletes are applied to the inserted data set. These queries also use MarkLogic features such as facets, pagination and bookmarks.

The benchmark test corpus uses the publicly available Wikipedia xml collection. Files are held on disk in zipped format, and the MarkLogic Content Pump (mlcp) is used to power an I/O intensive ingestion phase with the goal of generating the highest level of accuracy by forcing each device into steady-state availability.

The focus of the test is to look at overall latency from each evaluated storage solution across four areas of interest: journal writes (J-lat), save writes (S-lat), as well as merge read (MR-lat) and merge write latency (MW-lat).

As part of the benchmark, Mellanox Technologies, Inc.‘s IB interconnects were used to provide the highest performance and greatest network efficiency during testing. IB’s high-bandwidth, low-latency links allow for the greatest amount of data to be moved over the least number of links, allowing for the full system capabilities to be realized.

"Mellanox IB interconnects are designed to drive data volumes that are high enough to saturate the performance of tested platforms," said Gilad Shainer, VP of market development at Mellanox. "Only when all parts of a system are in balance, can the full performance potential of the overall system be achieved. Mellanox IB interconnects bring the data delivery connection up to par with MarkLogic’s flash-accelerated application performance, allowing for even higher performance in addition to increased infrastructure return on investment."