Elasticsearch Hosting on Your Own VPS
Full-text search and log analytics, with heap and retention under your control.
Size a server for Elasticsearch
Every stop is a real configuration you can deploy. Priced per 30 days, billed by the hour, exclusive of 18% GST.
What Elasticsearch is
Elasticsearch is a distributed search and analytics engine built on Lucene. It powers site search, product search, log analytics and anything else where "find documents matching this" needs to be fast over a large corpus.
It is a JVM application, and that shapes hosting more than anything else. Performance depends on heap sizing, and the guidance is specific: give the JVM no more than half the machine's memory, leaving the rest for the filesystem cache that Lucene relies on. Setting heap to the full machine size makes it slower, not faster, which catches people out.
The second thing to know is that a single node is a single point of failure by design. Elasticsearch is built to be clustered, and one node means no replica shards. That is perfectly reasonable for a search index you can rebuild from a source of truth; it is not reasonable if the index is the only copy of the data.
Why run Elasticsearch on a VPS
Heap and filesystem cache balance under your control, which is where search performance lives.
Retention is a disk decision, not a per-document bill.
Root access for the kernel settings it requires, such as vm.max_map_count.
A private network for cluster traffic if you grow to more than one node.
Elasticsearch features
Full-text search with analyzers and relevance tuning
Aggregations for analytics and faceting
Index lifecycle management for time-series data
REST API with client libraries for most languages
Snapshot and restore to local or object storage
Kibana alongside it for exploration
Recommended server
A starting point for Elasticsearch, not a hard floor — size it on the sliders.
| Memory | 8 GB minimum — half to JVM heap, half to filesystem cache |
|---|---|
| CPU | 4 vCPU; indexing and aggregation both use it |
| Storage | 80 GB and upward; indexes are larger than the source documents |
| Operating system | Ubuntu 22.04 LTS or AlmaLinux 9 |
How deployment works
Elasticsearch is installed with heap set to roughly half of system memory — the rest is left deliberately free for the filesystem cache Lucene depends on. vm.max_map_count is raised, which the service refuses to start without. It binds to a private interface with security enabled; an open Elasticsearch node is a well-known data-loss route.
What people run it for
Site and product search
Typo tolerance, faceting and relevance a SQL LIKE cannot approach.
Log analytics
Centralised logs with aggregation, retained as long as your disk allows.
Magento search
Magento 2 requires it, and running it beside the store keeps queries local.
Analytics over documents
Aggregations across large document sets without a warehouse.
Elasticsearch hosting questions
Related applications
Redis Hosting In-memory cache, sessions and queues, priced as memory rather than per gigabyte. 
Grafana Hosting Dashboards and alerting over your own metrics, without per-seat or per-metric billing. 
Magento Hosting Magento 2 on a server actually sized for it, with the search and cache services it requires. 
ClusterControl Hosting Deploy, monitor and fail over database clusters from one interface. 
Supabase Hosting Postgres with auth, storage and realtime APIs, on your own server. Deploy Elasticsearch on your own server
Root access, your choice of size, and no container between you and the application.