Speed Does matter - SingleStore is way forward Solution
December 16, 2020

Speed Does matter - SingleStore is way forward Solution

Anonymous | TrustRadius Reviewer
Score 8 out of 10
Vetted Review
Verified User
Review Source

Overall Satisfaction with SingleStore (formerly MemSQL)

Before having [SingleStore (formerly MemSQL)] in our organization we used to have SpiCA DB. The challenges being the performance of the DB tends to degrade with the increase in the OLTP significantly over the years. This has resulted in existing DB to perform slow and end customer experience [deteriorated]. Hence a tool was looked for that can improve the performance of the overall E2E flow of transactions and also be scalable to for high level of transactions processing. Currently in our organization [SingleStore] is used as backbone for all the batch data loads and real time messages storing in our line of business. Its success has resulted in pitching in voice to be implemented on other LOB as well.
  • First and foremost with the use of Aggregator master and leaf nodes the query processing time for DB are reduced significantly from hundred of msec to now just below 50msec.
  • Supports very high level of Real time transaction processing, for our project the OLTP via real time messages were ~500 to max of 800 TPS, the SingleStore was able to process those very efficiently with significant lower sever side utilization.
  • One of the most powerful tool for SingleSore monitoring is Planecache query, this helped in measuring performance of the stored procedures and evaluating the optimization required. The use was simple and helped analyst to provide the outcome in shortest interval of time.
  • Features related on [SingleStore (formerly MemSQL)] studio UI for monitoring needed some improvement when we used. The SingleStore studio support team helped us in resolving the issues with relation on running the different sets of indexing queries. Challenges here being the detailed logs were not available and we needed to connect with DBA for more detail drill down.
  • SingleStore can have some more improvement in setting up more examples on help documentation for the beginners and early learners.
  • Increased support of higher transactions by the components.
  • Value for money as more transaction per second can be achieved resulting in greater revenues.
  • Overnight batch processing time reduced from 8 hrs to 2hrs 10 min, resulting in less support team time.
The other tool was Oracle. The unique selling price for SingleStore was the efficiency when compared with set of data points. The other factor being the hardware infrastructure required. In case of setting up the Oracle setup the number of servers needed to process the volumes were considerable more than that used for setting leaf nodes and aggregators.
Lastly the maintenance activity for Oracle required more manpower compared to SingleStore setup with customization of the alerts and responses.
The SingleStore provide significant competitive advantage namely early launch to market, reduced batch process time compared to similar tool.
In our organization the AI component of SingleStore was extensively implemented.

SingleStore has capabilities to plug in into modern applications and workflows, such as AI workflows. It allows us to close the loop and automate the loop. Festure enables to make the system completely automatic.

We are currently testing technology internally that reliably allows us to get responses to a specific set of query types both from row and Columnstores. I was demo with the business sponsors, and they were quire excited.

Additionally SingleStore is particularly well-suited for use in speeding up the workflows that are needed for all kinds of AI and machine learning applications as well.

Our existing system was limited in terms of scalability due to high cost in investment for upgraded hardware. SingleStire provided efficient way to scale our relational database that supported more than 40 upstream components.
SingleStore is an converged database for operational analytics hence was used for fast data ingestion, low-latency queries, and elastic scaling with familiar relational SQL. MemSQL distributed database to easily capture, process, analyze, and act on data to thrive in today’s insight-driven.
Since SingleStore is a cloud-native distributed, highly-scalable, relational SQL database it is currently able to handle both OLTP and OLAP workloads in a single system in our organization. With changes in the business flows it very well fits with the direction of new applications changes to combine transactional and analytical (HTAP – Hybrid Transaction/Analytical Processing) requirements. Looking ahead for enterprise level we are standardizing container-based approaches to ensure that applications can consistently run and be managed identically, regardless of the underlying cloud infrastructure.

Do you think SingleStore (formerly MemSQL) delivers good value for the price?


Are you happy with SingleStore (formerly MemSQL)'s feature set?


Did SingleStore (formerly MemSQL) live up to sales and marketing promises?

I wasn't involved with the selection/purchase process

Did implementation of SingleStore (formerly MemSQL) go as expected?


Would you buy SingleStore (formerly MemSQL) again?


SingleStore [formerly (MemSQL)] is very well suited for places
- where the components are processing very large amounts of data and requires very low latency.
- Columnstore compression of data reduces the time to respond. Compression resulted in quick responses which are not achieved using the other DB tools.
- The concept of Rowstore and implementation on frequently used tables results in support of high OLTP.

Not suited/less appropriate
- The In-memory(Rowstore) and col-store does not share the same language compatibility. When required the transition form other table type more efforts are required.
- SingleStore DB (formerly MemSQL) connection between AWS cloud failed when partitioning is higher for data processing.
- Administration is sometime bit confusing when providing layered access to different teams.