{"id":6989,"date":"2023-11-15T08:00:53","date_gmt":"2023-11-15T16:00:53","guid":{"rendered":"https:\/\/devblogs.microsoft.com\/cosmosdb\/?p=6989"},"modified":"2023-11-13T10:11:05","modified_gmt":"2023-11-13T18:11:05","slug":"whats-new-for-azure-mi-for-cassandra-5-0-at-microsoft-ignite-2023-building-the-next-generation-of-ai-apps","status":"publish","type":"post","link":"https:\/\/devblogs.microsoft.com\/cosmosdb\/whats-new-for-azure-mi-for-cassandra-5-0-at-microsoft-ignite-2023-building-the-next-generation-of-ai-apps\/","title":{"rendered":"What\u2019s New for Azure Managed Instance for Apache Cassandra at Microsoft Ignite 2023: Building the Next Generation of AI Apps"},"content":{"rendered":"

\"Image<\/a><\/p>\n

 <\/p>\n

Get ready for Microsoft Ignite 2023, where we’ll be unveiling exciting new features and enhancements for Azure Managed Instance for Apache Cassandra<\/a>. With the introduction of vector search, turnkey major version upgrades, turnkey replication for multi-region clusters, turnkey hybrid cluster setup for live migrations, and L-series VM SKUs with write-through cache, developers can now handle major Cassandra upgrades seamlessly, deploy multi-region clusters effortlessly, carry out live migrations easily, and reduce tail latencies and improve performance. Join us at Microsoft Ignite 2023 to discover how these new features can help you build next-generation AI apps with Azure Managed Instance for Apache Cassandra<\/a>.<\/span>\u00a0<\/span><\/p>\n

 <\/p>\n

Vector Search for Generative AI Apps<\/span>\u00a0<\/span><\/h3>\n

Cassandra 5.0 is the next major version of Cassandra, bringing dynamic data masking and vector search as core new capabilities<\/a><\/span><\/b>. This preview version includes a limited version of vector search that customers can try out, and dynamic data masking<\/a><\/span>.<\/span>\u00a0<\/span>Vector search is facilitated by storage-attached indexes that represent data as vectors in a multi-dimensional space. This enables fast and accurate similarity-based searches, which can be used to build next-generation AI apps that use OpenAI and other LLMs.<\/span> Learn more in our blog on Vector Search<\/a>.<\/span><\/p>\n

 <\/p>\n

Turnkey Version Updates<\/span>\u00a0<\/span><\/h3>\n

In addition to vector search, we’re also introducing turnkey version upgrades. Previously, the only way to upgrade between major versions was to perform a data center migration. <\/span>Now, customers can perform in-place major version upgrades directly from the portal, or using Az CLI, Terraform, or ARM templates.<\/span><\/b><\/p>\n

 <\/p>\n

Streamline Deployment of Multi-Region Clusters with Turnkey Replication<\/span>\u00a0<\/span><\/h3>\n

We’ve made it even <\/span>easier to deploy multi-region clusters with turnkey replication<\/span><\/b>. Previously, setting up multi-region replication required several configurations. With this enhancement, deploying a multi-region account is much more straightforward.<\/span>\u00a0<\/span><\/p>\n

Live Migrations with Turnkey Hybrid Cluster Setup<\/span>\u00a0<\/span><\/h3>\n

Live migrations are also made easier with turnkey hybrid cluster setup. Previously, setting up a hybrid cluster required a complex and error-prone set of steps. <\/span>This feature introduces two enhancements to make migration more seamless: a command that creates a script to be run on the source cluster, and a new portal experience that shows progress on streaming during replication.<\/span><\/b> Learn more from our <\/span>docs page.<\/span><\/a>\u00a0<\/span><\/p>\n

 <\/p>\n

Write-Through Cache<\/span>\u00a0<\/span><\/h3>\n

Finally, <\/span>we released write-through caching by leveraging the L-series VM SKUs to reduce tail latencies and improve read performance intended for heavy workloads<\/span><\/b>. These SKUs have locally attached disks for fast IOPS and lower tail latency. However, when self-hosting these SKUs, the local disks are volatile and there is a possibility of data loss. This new SKU includes automated mirroring of the local disk data to P30 disks that are attached, providing a transparent way to resolve data durability issues.<\/span>\u00a0<\/span><\/p>\n

We’ve released a host of new features and enhancements to help developers build next-generation AI apps with vector search, handle major Cassandra upgrades seamlessly, deploy multi-region clusters effortlessly, carry out live migrations easily, and reduce tail latencies and improve performance with write-through cache. Discover more about these exciting new features at Microsoft Ignite 2023.<\/span>\u00a0<\/span><\/p>\n

 <\/p>\n

Learn more about Azure MI for Cassandra\u00a0<\/span>\u00a0<\/span><\/h3>\n