Sharding and Replication

13 posts · Jan 2026 – Sep 2026

What this trail covers

Meilisearch Cloud is evolving to support large SaaS platforms with millions of tenants requiring per-tenant search isolation. This involves distributed search capabilities like sharding and replication for horizontal scaling and high availability, and the introduction of serverless indexes. Serverless indexes will significantly reduce costs by moving inactive indexes to object storage and only spinning them up when queried, making per-tenant isolation affordable at scale and enabling a real free. The engine now includes replicated sharding, a step towards one-click replication in the Cloud UI.

All Meilisearch trails

Sep 2026How to diagnose search performance problems in B2B SaaSThis post provides a framework for diagnosing and troubleshooting search performance problems in B2B SaaS environments. It breaks down "fast search" into five dimensions for backend engineers: query latency, indexing latency, network latency, availability and recovery, and performance under load. It details diagnostic steps for each dimension, including the use of Meilisearch's performance trace, task queue inspection, batching strategies, and considerations for regional deployment and scaling. It explicitly mentions geo-replication and sharding as enterprise capabilities for addressing slow search for users in distant regions and slow search at scale, respectively.
Jun 2026CarJager: Millions of Vehicle Searches on MeilisearchThis post details CarJager's migration from Algolia to Meilisearch, highlighting the challenges faced with Algolia's sorting limits and pricing model. It emphasizes Meilisearch's advantages in API familiarity, lack of artificial sorting limits, autoscaling capabilities for traffic peaks, and a more aligned cost structure for managed instances. The migration was phased, starting with a low-stakes index and progressing to the core car listings, with Meilisearch's Solution Engineering team providing support. The results show improved developer experience, native sorting across the full catalog, seamless operations through autoscaling, and significant cost reduction.
Apr 2026Meilisearch Launch Week April 2026 recap: everything that shippedThis post announces the general availability of sharding and replication on Meilisearch Cloud. It details the use cases for horizontal scaling, high availability, and geographic replication. The composability of sharding and replication is highlighted, allowing for flexible configurations. The post also mentions that customers have been running production workloads on sharded Meilisearch for months, and this release brings that capability to all Cloud teams.
Apr 2026Scale without limits: introducing sharding & replication in Meilisearch CloudIntroduces sharding for horizontal scaling of large datasets and replication for high read availability and geo-distribution to Meilisearch Cloud. Sharding distributes data across nodes, with queries fanning out and results merging. Replication keeps multiple nodes in sync with the same complete dataset for failover. The post details how sharding and replication are composable building blocks, enabling configurations like sharding-only, replication-only, or a combination of both. It also outlines future roadmap items including optimal geo-replication with sharding and self-serve replication UI.
Apr 2026How Touchless powers instant, scalable vehicle search for large dealer groupsThis post details how Touchless uses Meilisearch to power instant, scalable vehicle search for large dealer groups. It highlights the switch from Algolia due to Meilisearch's API flexibility, integration with edge infrastructure (Cloudflare edge workers and web sockets), and cost-efficiency. The implementation involves a normalization pipeline, Sanity as a database, and dedicated Meilisearch indexes per dealer group, with new indexes spinning up in milliseconds. The system handles search, filtering, and infinite scroll for inventories of up to 9,000 vehicles across multiple locations, emphasizing the performance benefits of Meilisearch's architecture in a multi-tenant SaaS product.
Advertisement
Mar 2026The good, the bad, and the leaky: jemalloc, bumpalo, and mimalloc in MeilisearchThis post details the investigation and resolution of memory leaks in Meilisearch, primarily focusing on the interaction between different memory allocators (jemalloc, mimalloc, bumpalo) and the LMDB dependency. It highlights a leak related to `bumpalo::Vec::into_bump_slice` and a more significant issue stemming from the non-cooperation of LMDB's system allocator with Meilisearch's custom global allocator (mimalloc). The solution involved unifying the allocator by using jemalloc and later mimalloc v3 with its override feature, leading to significant memory usage reduction and performance improvements.
Mar 2026Meilisearch roadmap roundup: where we're headingThis post outlines Meilisearch's roadmap, detailing significant advancements in distributed search (sharding and replication) and the upcoming introduction of serverless indexes. It also describes plans to enhance hybrid search by simplifying embedding setup, introducing an AI gateway for caching and provider management, and developing proprietary models. The roadmap further includes transforming Meilisearch into a platform by exposing more engine capabilities through the Cloud dashboard and integrating advanced debugging and observability tools. Lastly, it details plans to scale RAG capabilities with engine improvements like parallel multi-search and simplified agent configuration, alongside developer-focused tools like a chat playground and UI starter kit.
Mar 2026Patching LMDB: how we made Meilisearch’s vector store 3× fasterThis post details a significant optimization to Meilisearch's vector store by patching LMDB to enable concurrent read transactions on uncommitted write transactions. This was achieved by implementing nested read-only transactions within a parent write transaction, allowing multiple threads to read uncommitted changes lazily on demand. This eliminates the need for a large, slow-to-initialize HashMap to store pointers, reduces cache pollution from full scans, and enables indexation to scale linearly with the number of CPUs, improving time complexity.
Mar 2026Meilisearch March updatesThis post announces the integration of replicated sharding directly into the Meilisearch engine (v1.37.0), a significant step towards distributed, resilient search. This feature is currently available under the Enterprise edition, with one-click setup in the Cloud UI planned for the future. Additionally, it mentions improvements to the HNSW-backed vector store (Hannoy) and faster embedding updates in v1.38.0, along with fixes for remote embedders and task deletion.
Feb 2026How Scenario scales asset discovery with MeilisearchThis post details how Scenario scaled their asset discovery with Meilisearch, moving from OpenSearch on AWS. It highlights the challenges of performance at scale, cost/operational overhead, and growing user expectations. Scenario chose Meilisearch for its developer-first experience, enterprise readiness, and hybrid search capabilities, enabling them to combine proprietary embeddings with keyword and multi-modal search. The implementation includes continuous indexing, custom embeddings, and vector search integration for an "Exploration" feature, resulting in dramatic reduction in support load, exceptional scale handling (35M+ searches/month, 80% vector), and productivity gains for their engineering team.
Jan 2026Meilisearch January updatesIntroduced dynamic shard addition for horizontal scaling and flexibility in splitting indices across multiple instances. This allows scaling without re-architecting and maintains fast speeds via federated search. Also noted that replicated sharding is coming soon.