Elasticsearch hosting
Elasticsearch is a powerful, distributed search and analytics engine. It enables real-time full-text search, scalable data analysis, and flexible filtering capabilities. Whether for websites, applications, or business intelligence – Elasticsearch delivers fast and precise results even with large volumes of data.
- Lightning-fast full-text searches (near real-time)
- Horizontal scaling & high availability
- Real-time aggregations & faceting
- GDPR compliant
All-round carefree package
Our managed Elasticsearch hosting packages
Choose the package that matches your requirements – our hosting solutions offer different performance levels with exactly the features and services your project needs.
Elasticsearch hosting Standard
Standard
69 €/month
138 € setup
- Own VM
OpenStack Public Cloud - 2 cores, 4 GB RAM
Shared resources - 50 GB disk
RAID 10 NVMe SSD - 250 Mbit/s
Unlimited traffic - Guaranteed availability
99% at application level
99.9% at network level - Data center Germany
ISO 27001 certified
- Elasticsearch instance: 1
- Update management
- Backup management
Daily full backups, 14 days retention - Active monitoring
- SLA / DPA
Elasticsearch hosting Premium
Premium
149 €/month
298 € setup
- Own VM
VMware vSphere Private Cloud - 2 cores, 4 GB RAM
Dedicated resources - 50 GB disk
Redundant SSD storage - 10 Gbit/s max
Unlimited traffic - Guaranteed availability
99.5% at application level
99.95% at network level - Data center France
ISO 27001, SOC 1, SOC 2 certified
- Elasticsearch instance: 1
- Update management
- Backup management
Daily full backups, 14 days retention - Active monitoring
- SLA / DPA
Elasticsearch hosting Enterprise
EnterpriseRequest a quote
You get exactly the environment your project requires. Briefly tell us what you have in mind – we will design the appropriate architecture and get back to you with a concrete proposal.
All prices plus statutory VAT.
Solutions for Agencies and developer teams · Public bodies
Managed Elasticsearch hosting details
Full feature list
| Item | Elasticsearch Hosting Standard | Elasticsearch Hosting Premium | Elasticsearch Hosting Enterprise |
|---|---|---|---|
| Prices | 69 € monthly 138 € setup | 149 € monthly 298 € setup | On request |
| Application | |||
| Elasticsearch instance | 1 | 1 | |
| Update management | ✔ | ✔ | |
| Backup management SaaS Web creates full backups of the application every day and keeps them for 7 or 14 days depending on the plan. | daily | daily | |
| Backup retention | 14 days | 14 days | |
| Monitoring With active monitoring from SaaS Web the core parameters of the server are checked every minute. If thresholds are reached, we step in to get things back to normal as quickly as possible. | ✔ | ✔ | |
| Hosting | |||
| Platform | Own VM | Own VM | The hosting setup can be customized individually:
|
| Infrastructure | OpenStack Public Cloud | VMware vSphere Private Cloud | |
| Performance | 2 cores, 4 GB RAM | 2 cores, 4 GB RAM | |
| Resources | Shared resources | Dedicated resources | |
| Disk | 50 GB | 50 GB | |
| Storage type | RAID 10 NVMe SSD | Redundant SSD storage | |
| Internet connectivity | 250 Mbit/s | 10 Gbit/s max | |
| Traffic | Unlimited | Unlimited | |
| Guaranteed availability | 99% at application level 99.9% at network level | 99.5% at application level 99.95% at network level | |
| Support | Ticket system, email, phone | Ticket system, email, phone | |
| SLA | ✔ | ✔ | |
| Data center | DE (Limburg) | FR (Roubaix) | |
| Certification | ISO 27001 | ISO 27001, SOC 1, SOC 2 | |
| Anwendung | |||
| Domains On request SaaS Web registers a domain free of charge and connects the booked service to that domain. | At additional cost | At additional cost | The application can be configured as desired:
|
| Order | Order | Enquire | |
More about Elasticsearch
Full-text search & analyzers
Elasticsearch breaks down text via tokenizers and refines them with analyzers (stemming, lowercasing, stopwords, synonyms, ICU/language packs). This ensures that “running” and “run,” brand names in different spellings, or terms with umlauts match consistently. You can use different analyzers per field (e.g., search_analyzer ≠ analyzer) to fine-tune input vs. query behavior.
Query DSL (Bool/Filter/Range/Term/Match)
The JSON-based DSL combines structured filters (term/range/exists) with full-text queries (match/multi_match) in bool clauses (must, filter, should, must_not). Filters are cacheable (fast), while match queries contribute to scoring. This allows for the construction of complex, high-performance search logic – from “category + price range + free text + availability” to special cases with script queries.
Relevance & scoring
The standard is BM25, supplemented by boosts (fields, terms, tie_breaker), function_score (e.g., popularity, freshness via gauss/exp decay), rescore for top-N refinement, as well as min_should_match and boost_mode. This allows you to prioritize current, popular, or exact matches without losing weaker ones.
Aggregations (Bucket/Metrics/Pipeline)
Buckets (terms, date_histogram, range) group documents; metrics (avg, sum, percentiles, cardinality) provide key figures; pipeline aggregations (moving_avg, derivatives, bucket_script) calculate trends from intermediate results. Everything can be combined with the search in a single request (“search-and-analyze”).
Autocomplete & suggestions
The completion suggester maintains predefined phrases in the FST index for extremely fast prefix completion. Term and phrase suggesters suggest corrections for typos; fuzzy match tolerates deviations (Levenshtein distance). For search boxes, this means: if you type “iph,” you immediately get “iPhone 15 Pro …”.
Index management & mappings
Mappings define field types (keyword, text, numeric, date, geo_point, nested) and analyzers. With index templates and aliases, you control schemas and zero-downtime reindexing (switching alias from old → new). ILM (Index Lifecycle Management) automates phases: hot (fast), warm (compressed), cold (economical), frozen/delete.
Scaling & high availability
Sharding distributes data horizontally across nodes, while replicas ensure fail-safety and parallel reading. Cluster balancing moves shards automatically, and rolling upgrades maintain high availability. You grow linearly: more data, more nodes.
Ingest pipelines
Before indexing, documents pass through processors: grok (log parsing), geoip, date, set, rename, remove, script, attachment (Base64/PDF/Office extraction). This allows you to move ETL logic to the edge of the cluster without separate preprocessing.
Geo & time series functions
geo_point and geo_shape enable distance, bounding box, polygon, and within queries; sort by _geo_distance for proximity. For time series: date_histogram, downsampling (via rollup/transform), moving_fn in pipelines. This is optimal for maps, routes, proximity searches, and metric/log time series.
Security & access control
TLS/HTTPS secures transport, while API keys and users/roles (RBAC) limit access. Field/document-level security hides sensitive fields/documents; auditing logs access. With SSO/OIDC/SAML, you integrate corporate identities.
Frequently asked questions
Can I move my existing Elasticsearch server to SaaS Web?
Yes. We handle migrations via snapshot/restore or “reindex from remote” – depending on the version, data volume, and downtime tolerance. Frequently, we first create a snapshot (repository such as S3/Swift/NFS), import it into a staging cluster on our end, check mappings, plugins, and security settings (RBAC, TLS, API keys), and then switch live via index alias or a read-only phase with a brief interruption. Alternatively, we migrate with virtually no downtime by indexing in parallel and performing the switchover in a short maintenance window. We also support version upgrades (e.g., 7 → 8) including ILM/SLM adjustments, Kibana spaces/dashboards, and monitoring.
When is Elasticsearch more useful than the full-text search of a traditional database?
As soon as relevance ranking, typo tolerance, synonyms, multilingual analyzers, autocomplete, and extensive aggregations are required – and all that with millions of documents at low latencies. Databases deliver solid “contains” matches but quickly reach their limits with complex ranking, faceting, and scaling. If you need a search-centric UX (shop, help center, log/metric analytics), Elasticsearch is the better choice; for very small data volumes and simple requirements, DB search may be sufficient.
How do I model fields and indices correctly so that search and filters remain precise and fast?
Free text should be included in the index as text with a suitable language analyzer, while exact filters/sorting/aggregations run via keyword. Often, both are combined in the same field via fields (e.g., title as text plus title.keyword). Arrays of objects that logically belong together (variants, attributes) should be modeled as nested so that queries do not produce cross-matches. For autocomplete and error tolerance, specialized analyzers (e.g., edge_ngram) and suggesters are useful; different analyzers for index and search time (analyzer vs. search_analyzer) help balance precision and recall. Ideally, you should provide indices with templates, aliases, and ILM rules so that rotation, compression, and retention are properly automated.
Partner Program
With our partner program you benefit from attractive discounts and low rates for development work – with no long-term commitment and no complicated requirements.
- SaaS Web Coins: up to 20% discount on hosting
- Lower hourly rates for development work
- WordPress plans: data protection package at half price
Developer friendly
Our services are interesting for developers too: we address their particular needs and deliberately support modern ways of working and individual setups.
- 1 month free
You build the website for your client, who takes over the hosting when it goes live. - SSH (SSH key access), SFTP access, Git integration
- Custom setup options
Testimonials
See how other customers use our solutions and what their experience has been.
SaaS Web provides a steady pulse in the background for us: Migrations, updates, peaks – everything runs smoothly. Custom setups are implemented pragmatically, releases happen without downtime, and for questions, you reach real engineers instead of getting stuck in ticket loops. This way, our energy remains focused on content and users – not on hosting.
STJA
With SaaS Web as our partner, our Managed Kirby platform runs smoothly: fast provisioning, professional customization for individual setups, and a genuine understanding of our requirements—allowing us to fully focus on creativity, development, and client results, while SaaS Web reliably takes care of the rest.
Bold and Bright LLC
SaaSWeb impresses with reliable SaaS hosting as well as the professional development of Craft CMS websites and applications. Projects are implemented efficiently and managed technically at the highest level.
Poli Marketing GmbH
The name says it all: SaaS Web!
Prompt support and our hosting packages are always up to date, whether for TYPO3, WordPress, or VM. This is very important for us as an agency—for all our projects, we receive well-thought-out solution proposals with alternatives.
T2U GmbH
Our customers
Many companies already rely on our solutions. Here is a selection of our customers.
With SaaS Web as our partner, our Managed Kirby platform runs smoothly: fast provisioning, professional customization for individual setups, and a genuine understanding of our requirements—allowing us to fully focus on creativity, development, and client results, while SaaS Web reliably takes care of the rest.
