Time series database firm EraDB launches Erasearch log management

Time series database firm EraDB launches Erasearch log management

Time series database vendor EraDB on Feb. 10 launched its first commercially supported product with the general availability of EraSearch.

A time series database is designed to store and organize data as it occurs in a time sequence. Multiple time series databases are available in the market today including the Amazon Timestream service that can be used for such applications as financial services, fraud detection and log management.

Rather than providing a general-purpose time series database, Seattle based startup EraDB is launching the Erasearch platform as a focused product to provide log management.

Erasearch is compatible with the Elasticsearch log search technology and integrates proprietary capabilities for data management, scalability and performance.

Mike Leone, senior analyst at Enterprise Strategy Group, said that by decoupling storage and compute, EraSearch gives organizations a way to reduce the size of their infrastructure to an optimal level. Another aspect of the Eraserach platform Leone said is valuable to organizations is the idea of auto-rebalancing.

“Too often organizations experience performance penalties due to ineffective or untimely rebalancing of data, creating hot and cold spots,” Leone said. “EraSearch is leveraging AI to automatically rebalance shards of data to ensure data is where it needs to be to deliver the expected levels of performance.”

How EraDB expanded its time series database to Erasearch

EraDB co-founder and CEO Todd Persen is no stranger to the world of time series databases. Persen was formerly a co-founder and CTO at InfluxData, another time series database vendor.

Persen explained that in his view, time series databases have sometimes struggled with high cardinality data as well as the challenge of managing large volumes of data, which are problems he said he’s looking to attack with EraDB.

The vendor got started building out the EraDB database in late 2019 and began to engage with potential customers to try and find a market fit.

One of the things Persen said that he kept hearing from prospects was that they were doing log management with Elasticsearch but were having trouble scaling.

Persen and his team started working in 2020 with a large company that was willing to do a pilot engagement and in concert with the user, EraDB began building out a purpose-built offering for log management, built on top of that EraDB time series database infrastructure.

“At the end of the day, you know, logs are a time series, they’ve all got a timestamp and logs are largely structured,” Persen said.

Dynamic synchronized caching in Erasearch

Among the key features in Erasearch is a capability the company refers to as dynamic synchronized caching.

Persen explained that in the cloud, when a technology platform decouples storage, and compute, it often means putting the data in a cloud storage service such as Amazon S3. The challenge according to Persen, is that while S3 is great for low-cost storage, it’s not as effective for performance retrieval.

“For log management, people expect really relatively fast interactive type queries,” Persen said.

With dynamic synchronized caching, EraDB has built a caching tier that can actually optimize the performance of log data. The cache can be configured to optimize data for a given time period that reflects what the system determines to be the most likely data to be requested.

What’s next

While log management is the initial use case for the EraDB time series database technology, it’s not the only market the vendor will be going after, Persen said. Other potential areas that EraDB  is looking at include anomaly detection for security use cases.

“We will continue to build and refine the Erasearch product,” Persen said. “We’ll also take our perspectives on time series database as a technology for solving other categories of data problems and start to build vertically on top of that.”

ESG is a division of TechTarget

 

 

Source link