Heimdall Data – Reduce SQL Server costs

The Heimdall Data software agent automatically caches the data without disrupting the current infrastructure.

Caching can optimize application response time and reduce database licensing costs. However, this is not an easy task: due to the need for manual configuration (i.e. to select data for caching), it requires additional resources and increases the risks. In this article I will discuss how the Heimdall Data tool automatically caches the data without disrupting the current infrastructure.

Why is it caching

• Studies have shown that page loading times are reduced by 20% and conversion rates are increased by the same amount.
• Performance is improved and application response time is shortened.
• Database licensing costs are reduced.
• Increases the horizontal scaling of the database.

What to cache

A typical list of control questions when organizing caching is as follows:

• Which data can be cached safely?
• Will the caching yield a performance gain?
• How can I keep my cache up-to-date?

Suitability for caching depends on relevance and synchronization. What are the guarantees that a record made on one host will be immediately updated on the other? Without this guarantee, caching is risky and is likely to disrupt application performance. A caching solution must adapt to changes in traffic to optimize performance.

Application layer caching

SQL Server provides basic caching (such as plan and buffer pool queries), but does not solve the problem of network latency when queries arrive in the database and lead to losses due to repeated transmission on the application server.

Some organizations deploy additional caches on the application side, such as Redis, Amazon ElastiCache and Hazelcast: These are SQL caches in front of the database designed to improve application performance and reduce database processing costs. They reduce memory constraints for the buffer pool and improve the efficiency of queries processed by the database server.

However, application-level caching is performed manually, so errors are possible. Application owners still need to determine which data to store in the cache, often track changes in the data structure and make appropriate adjustments. What if there was an automatic caching solution?

Automatic caching with Heimdall Data

Technical units spend up to 20% of their resources on building and maintaining a caching subsystem. This operating cost can be avoided with a Heimdall Data caching solution:

• • Intelligent auto caching that does not require changes to applications or SQL Server.
  • Machine learning algorithms perform the caching only if they manage to gain performance gains and automatically perform the cancellations.
  • Intelligent caching on two levels (see figure):

• – at the user’s choice (for example, Amazon ElastiCache, Redis).
  – in memory;

Drawing. Heimdall Data Caching Architecture

Heimdall Data is a software agent that works with any application, including third party applications, as no changes in the software code are required. It is enough to change the network settings to connect the application to the Heimdall Data mediating server, which connects to the database on behalf of the application.

Today’s caching solutions lack invalidation logic except lifespan or require complex settings. This is a risky, time-consuming, manual process. Heimdall intelligently caches and cancels data, eliminating the risk and need for guessing.

Getting started

When setting up applications to use a Heimdall SQL Server Reseller, no changes are made to the application except for changes in the URL and the “node + port” pair of the database; no changes are required in either the program code or the database system (screen 1). As a result, query traffic and server performance level are displayed on the monitoring panel on screen 2. Note that the average request time for cached traffic is 50 microseconds compared to 1000 microseconds from the database. By deploying the cache from this application, we get a performance gain of more than 20 times at the data level!

Screen 1. Heimdall Data installation wizard

Screen 2. Heimdall Data Monitoring Panel

Other statistics include cache hits and request traffic for each server. With a cache hit rate of 90%, the database load is significantly reduced, allowing it to serve more users in one database infrastructure.

Thus, Heimdall Data automates caching for SQL Server. Users will get a performance gain and reduce the need to purchase additional SQL Server licenses. By reducing the load on the database, fewer SQL Server instances are required. A free trial version of Heimdall can be downloaded from the Heimdall Data website.

Heimdall Data installation video