Dynatrace adds price tier, data pipeline amid cost crunch

Dynatrace pipeline, data observability adapt to growth

Dynatrace also unveiled plans this week to roll out a data pipeline service over the next 90 days that can shift some data processing away from its back-end platform. This new OpenPipeline feature is similar to what's offered by other observability data pipeline vendors such as Cribl, CloudFabrix, Mezmo and Calyptia (now owned by Chronosphere), and has a similar purpose: reducing data transfer, ingestion and storage costs by processing more data at the source.

OpenPipeline will initially support logs and business metrics, but it could potentially be used in security to detect unusual spikes or drops in data generated by applications and alert IT teams to these changes, which could be indicative of an incident, said Alois Reitbauer, chief technology strategist at Dynatrace.

"If, for some reason, your reporting density massively changed, [where] usually you have 10,000 log entries per minute and suddenly it's down to five … and suddenly all the intrusions in your log data are gone, [we could say], 'Your data looks kind of odd,'" Reitbauer said. "It looks like you … got hacked. But if they would only analyze the data [itself], it would say, 'Your system is working perfectly.'"

Dynatrace will also expand its data observability feature within the next 90 days to pre-analyze and cleanse data from third-party systems before ingestion into its platform. The shift will expand the data sources users can feed into its multi-modal AI analytics and IT automation systems.

This is also a feature embraced by log management vendors such as Sumo Logic, as well as observability data pipeline vendors, which support a variety of data sources. But in the case of newer startups such as Flip AI, data ingestion into a separate back-end platform isn't required, Gohring said.

"Flip AI -- which has built a domain-specific LLM [large language model] for pulling out root-cause analysis from observability data -- doesn't do any data collection or storage. It runs on whatever you want," she said. "That makes you start thinking about data collection, storage and analysis as three distinct requirements that don't have to come from the same vendor."

Dynatrace will also consider leaving data in multiple locations while doing analysis and argued that its OpenPipeline can better support massive, petabyte-level scale than its smaller competitors, according to Reitbauer.

Dynatrace and other vendors must continue adding value to back-end platforms to justify data ingest costs as open source projects such as OpenTelemetry perform the same function for free, Gohring said.

"The only way that the more expensive vendors are going to continue to grow is if they offer something more valuable than cheap data collection," she said. "They have to continue to build features and aptly describe their value in helping customers quickly and easily understand what's happening in their applications and how to fix any issues."

Dynatrace touts BizDevOps expertise, folds in Runecast

To that end, Dynatrace updates this week also included fresh support for monitoring generative AI applications and reporting on their business value. Similar to updates from competitors Datadog and New Relic, Dynatrace can monitor LLMs, the tokens they use to analyze data, their underlying data repositories and the applications that integrate with them.

Dynatrace plans to differentiate from competitors by offering business insights on AI apps so that companies can decide if the resources it takes to support AI are justified by their return on investment, Reitbauer said.

"Every observability and monitoring vendor has AI monitoring," he said. "What we're focused on is linking this to the actual business case [for AI] and end-user requests."

New Relic also expanded support for business data last week by making its open source Pathpoint project officially supported in its New Relic One platform, but Pathpoint requires a manual process to set up business data collection. Dynatrace can automatically detect and correlate business metrics out of existing ingest streams, according to Reitbauer.

Meanwhile, Dynatrace acquired AI-driven security and compliance vendor Runecast this week for an undisclosed sum, expanding its platform features into the security posture management and cloud-native application protection platform realms.

While such platform expansion will likely further increase data ingestion requirements, Dynatrace's strength in application security tools along with observability could make it worth the investment for IT organizations that want to centralize security and infrastructure observability, said IDC analyst Katie Norton.

"This acquisition of Runecast is probably going to give Dynatrace quite a bit of additional capability, especially in vulnerability management," Norton said. "And there is a lot of shared data between security and observability."

In fact, in responses to a soon-to-be published IDC 2024 DevOps survey, 30.2% of 311 enterprise IT pros surveyed reported that they "have a shared model for security and observability, including tooling, teams and data sources." Another 42.1% said they share some of those resources, and 17.7% said they plan to move to a shared model.

The security and observability combination and a new Grail back-end architecture are more likely to set Dynatrace apart from competitors in AI monitoring than business analytics and FinOps features, said Andy Thurai, an analyst at Constellation Research.

"[Dynatrace's] Grail makes it easy to store structured and unstructured data together in an index-free, schema-free data lake, and the remediation workflow with Davis CoPilot is somewhat unique and could be a potential differentiator," he said.

Editor's note: Updated Feb. 1, 2024 to include Foundation & Discovery mode pricing.

Beth Pariseau, senior news writer at TechTarget, is an award-winning veteran of IT journalism. She can be reached at [email protected] or on X, formerly known as Twitter, @PariseauTT.