Scale out on demand.
Only pay for what you use.

The free trial is up to 14 days and includes $300 in credits to get started.

The free trial ends when you have either consumed $300 in credits or 14 days have elapsed. At that time, you can move to the free tier which is a shared pool of resources that can handle peak streaming rate of 50 KiB/s and up to 100 GB of storage. You can also swipe your credit card and use any virtual instance in the developer or standard editions. Contact us to use the premium edition.

Yes, Rockset offers a free tier as part of its developer. The free tier is a shared pool of resources that can support up to 2 GB of storage.

A virtual instance is a distributed set of compute and memory resources used for streaming ingestion and queries. With Rockset’s separation of compute-storage, you can spin up or down Virtual Instances on demand or configure auto-scaling. Virtual instances are separate from data storage.
Virtual instances come in different t-shirt sizes- Xsmall, Small, Medium, Large, etc.- with the amount of compute and memory scaling linearly as you increase size (ie: Small has double the compute and memory of XSmall).

Instance classes have different mixes of memory and compute. The general purpose instance class balances compute and memory resources and can be used for various use cases. The memory-optimized instance class provides double the amount of memory compared to the general purpose instance class. We recommend starting with the general purpose instance class for performance testing.

A virtual instance is a distributed set of compute and memory resources used for streaming ingestion and queries. A shared virtual instance shares a global resource pool and can see bursty performance. A dedicated virtual instance is a set of isolated resources for your account for predictable performance.

Rockset supports data ingestion from streams, databases, data lakes and data warehouses using built-in connectors or the Write API. Data is ingested into a collection, or a table in the relational world, that can be joined with data from other collections and sources. Rockset offers a number of ingestion transformations using SQL, including rollups or pre-aggregations.
Rockset streaming data ingestion uses compute and memory resources from Virtual Instances. For some data sources, Rockset offers bulk ingestion which is automatically triggered during an initial data load. With bulk ingestion , multiple virtual machines are spun up to quickly ingest data and then spun down when ingestion is complete. This makes the initial data load very fast and efficient.

Rockset is offered as a fully-managed cloud service and encrypts data in flight and at rest. It offers a number of security features related to authentication, authorization, data segregation, networking, encryption and logging that are outlined in Rockset’s documentation. Rockset is also compliant with SOC 2 Type II, GDPR, CCPA and HIPAA.

The Rockset platform is built to take full advantage of cloud infrastructure services and automated system healing using Kubernetes. Rockset has tiered storage and stores data in hot storage for performance and object storage for durability. Its virtual instances can easily recover from one or multiple hardware failures with little to no impact on requests to the Rockset service. Learn more about how Rockset designs for high availability and disaster recovery in this guide.

Rockset autoscales Virtual Instances up and down based on CPU utilization thresholds. This enables you to meet performance requirements without paying for unused compute resources.

Rockset separates storage from compute. As a result, all workloads can access the same shared, real-time data without any data duplication.
There are several methods available to further reduce the storage costs in Rockset including the following:

• Retention: Set a retention duration for a collection to automatically drop older data
• Compression: Compress data to reduce storage size using LZ4 and Zstandard
• Real-time aggregations: Rollup data or pre-aggregate data at ingest time, combining multiple documents into one

Rockset has a tiered pricing model to offer more cost efficiency as storage sizes scale. The pricing tiers range from: 0-100 GB, 100 GB - 1 TB, 1 TB- 100 TB and 100+ TB. A user pays per-tier pricing for their respective storage range. For example, a user with data stored in the AWS us-west-2 region with 15 TB of storage will pay $0.25 per GB-month for the first 100GB,  $0.20 per GB-month for the next 900 GB, and $0.16/GB-month for the remaining 14 TB.

Rockset separates compute from compute. As a result, you can spin up and down isolated virtual instances for ingestion and queries. This makes it cost-effective to run multiple workloads, multi-tenant applications, dev/test environments and more on Rockset.
There are several methods available to reduce compute costs in Rockset including the following:

• Autosuspend virtual instances: Suspend a virtual instance after a period of inactivity ranging from 15 minutes to 12 hours
• Autoscale virtual instances: Auto-scale a Virtual Instance based on CPU utilization
• Scale down virtual instances: Scale down virtual instances with the click of a button or API call
• Microbatching: Load data in scheduled time internals to reduce the time a virtual instance runs

Yes, Rockset provides basic support that includes access to email and livechat in the product for help. Rockset also has business support and priority support outlined in documentation.