How to Navigate the Costs of Legacy SIEMS with Snowflake

Legacy security information and event management (SIEM) solutions, like Splunk, are powerful tools for managing and analyzing machine-generated data. They have become indispensable for organizations worldwide, particularly for security teams. But as much as security operation center (SOC) analysts have come to rely on solutions like Splunk, there is one complaint that comes up for some: Costs can quickly add up. 

The issue centers around their volume-based pricing model. This model can force security teams to make difficult decisions on what data to ingest. There are a number of online threads — see here, here and here just to link to a few — dedicated to how best to control costs, while limiting how much an organization has to compromise its security. But what if security teams didn’t have to make tradeoffs?

This blog post explores how Snowflake can help with this challenge. Let’s start with five cost factors organizations need to consider with their legacy SIEM solution and how Snowflake can help.

Legacy SIEM cost factors to keep in mind

  1. Data ingestion: Traditional SIEMs often impose limits to data ingestion and data retention. Snowflake allows security teams to store all their data in a single platform and maintain it all in a readily accessible state, with virtually unlimited cloud data storage capacity.

Now there are a few ways to ingest data into Snowflake. Security sources can be ingested directly through native means such as streaming, stages, syslog, native connectors or secure data sharing. Snowflake’s Snowpipe service helps bring in new data easily, at a price that is tailored to an organization’s needs. The most common method is Snowpipe auto ingest, which works for security teams who regularly ingest machine data. But this method isn’t for everyone because loading small amounts of data slowly or many small files can cost more than other options.

Snowpipe Streaming is another method that can save security teams money. With Snowpipe Streaming there’s no need to prepare files before loading, making the cost of getting data more predictable. 

Security teams can also reduce their costs by loading certain datasets in batches instead of continuously. For example, they could load a lot of data that isn’t needed for instant detection three times a day instead of constantly streaming that data, which can lead to more significant savings.

  1. Data retention: Many legacy SIEMS delete activity logs, transaction records, and other details from their systems after a few days, weeks or months. With Snowflake, security teams don’t have to work around these data retention windows. Instead, all data is always accessible for analysis, which simplifies cost planning and the data management strategy. It also provides more reliable generation of key security metrics such as visibility coverage, SLA performance, mean time to detect (MTTD) and mean-time-to-respond (MTTR). Snowflake also helps security teams save time by automatically compressing and encrypting the data, making it ready to query.
  1. Detection and investigation processing: Security teams depend on detection rules to find important events automatically. These rules need computing power to analyze data and spot attacks. In the cloud, computing can be measured in various ways, like bytes scanned or CPU cycles. This affects how much it costs and how predictable the costs are for processing detections. While computing costs might not have been a concern with fixed hardware in the past, it’s a whole new game in the cloud. 

For security teams, investigations require computational power to analyze collected data similar to running detections. Some solutions utilize different engines, such as stream or batch processing, for detections and investigations, while others employ the same engine for both tasks. Snowflake helps security teams understand how the query engine functions at a basic level, which helps them effectively plan for the cost estimates of their investigations.

Moving away from volume ingest-based pricing

A traditional SIEM typically manages all the data ingestion, transformation, detection and investigation processing for security teams. While out-of-the-box connectors and normalization can be useful, customers end up paying more by the nature of legacy SIEMs that use ingest volume-based pricing models. 

It’s important here to understand how this pricing model works. Ingest volume-based pricing can vary among the different legacy SIEM vendors but the basic principle remains the same: the more data security teams send to the SIEM for analysis, the higher the cost.

By moving away from traditional volume-based pricing models, security teams can gain more control of what logs they have access to and how much they are spending. A consumption-based pricing model, like Snowflake’s, allows security teams to have all the data on hand while paying for only the compute resources they use, making security more cost-effective. Snowflake’s pricing model is designed to offer flexibility and scalability, giving security teams the ability to only pay for the resources they use without being tied to long-term contracts or upfront commitments. 

How Snowflake Works

An open-architecture deployment with a modern security data lake, and best-of-breed applications from Snowflake, can keep costs down while improving an organization’s security posture. A security data lake eliminates data silos by removing limits on ingest and retention. Organizations can use a security data lake to scale resources up and down automatically and only pay for the resources they use — potentially controlling their costs without compromising their security. 

Security data lakes can also help analysts apply complex detection logic and security policies to log data and security tool output. Security analysts can quickly join security logs with contextual data sets, such as asset inventory, user details, configuration details, and other information, to eliminate would-be false positives, and identify stealthy threats. 

The value proposition is clear: organizations can consolidate their security data affordably and gain the flexibility to query that data at any time. Snowflake empowers organizations to make data-driven choices for long-term gain. We’ll dive into some customer success stories to show the potential of this approach.

Real customer success stories

If done right, Snowflake customers can experience remarkable cost savings. Let’s take a closer look at some notable success stories across various industries.

At Comcast, Snowflake’s security data lake is now an integral component of their security data fabric. Instead of employees managing on-premises infrastructure, the Comcast security data lake built on Snowflake’s elastic engine in the cloud stores over 10 petabytes (PBs) of data with hot retention for over a year, saving millions of dollars. Automated sweeps of over 50,000 indicators of compromise (IOCs) across the 10-PB security data lake can now be completed in under 30 minutes.

Guild Education can claim “up to 50% cost savings” working with Snowflake and is just one example that highlights the potentially significant financial benefits organizations can unlock with the Snowflake Data Cloud.

By adopting Snowflake as its data lake for security events, corporate travel management company Navan achieved a best-of-breed security architecture that is both cost-efficient and cutting-edge. The results are impressive:

  • Over 70% cost savings by adopting a modern SIEM-less architecture
  • 15K+ hours saved in 8 months
  • 4x improvements in MITRE ATT&CK coverage in 8 months

Ready to witness the transformative power of Snowflake? 

Watch our demo and discover how you can revolutionize your data management strategy, unlock substantial cost savings, and propel your organization into a new era of efficiency and innovation. Learn how you can augment your Splunk strategy with Snowflake today. 

The post How to Navigate the Costs of Legacy SIEMS with Snowflake appeared first on Snowflake.

LATEST ARTICLE

See Our Latest

Blog Posts

admin June 12th, 2024

Bringing machine learning (ML) models into production is often hindered by fragmented MLOps processes that are difficult to scale with […]

admin June 12th, 2024

Discovering and surfacing telemetry traditionally can be a tedious and challenging process, especially when it comes to pinpointing specific issues […]

admin June 12th, 2024

Today’s data-driven world requires an agile approach. Modern data teams are constantly under pressure to deliver innovative solutions faster than […]