developers
company
blog contact
the full power of the deepstreamHub platform - on-premise or as private cloud
designed to work together as well as on their own and integrate with any popular or in-house systems
scaling connection concurrency, authentication and multi-protocol
Gateways allow hundreds of thousands of users to connect to the same deepstream. Multiple Gateways can front a single deepstream node and can be deployed as internet facing endpoints in a DMZ or fronted by load balancers or reverse proxies.
Gateways keep load off the deepstream server by handling authentication and permissions and integrate with SSO mechanisms and authentication endpoints such as LDAP/Active Directory, Kerberos or JAAS to make sure that only verified messages are able to enter your network.
Gateways also provide endpoints protocols such as MQTT, HTTP, STOMP, GraphQL or binary data allowing a host of devices and programming languages to interact with deepstream resources.
scaling connection concurrency, authentication and multi-protocol
deepstream servers are the realtime data distribution and synchronisation backbone of deepstreamHub enterprise. They manage persistence, storage access, horizontal data replication, remote procedure call routing, version management and a host of other features.
dsh deepstream is based on the deepstream.io open source server that has become the de-facto standard for scalable, webfacing realtime applications and caters to a wide spectrum of enterprise use cases. It does boast higher performance, numerous plugins to integrate with popular enterprise systems such as SSO, databases or monitoring endpoints and most importantly a decentralized clustering mechanism allowing it to scale to millions of concurrent connections and billions of messages.
deepstream is maintained by a large and dedicated community that’s constantly extending the server itself and its ecosystem of connectors, integrations, tools and plugins.
deepstream overviewefficient cluster management, failover and load balancing
Orchestrators manage deepstream and gateways clusters in the most efficient way, reconnecting, starting and stopping ressources as requirements change and directing incoming client connections to the most suitable endpoint based on load, network proximity and subscription topology. This solves the infinite scale loadbalancing problem of concurrent connection and always ensures optimal uptime, high availability and low network latency for your end-users.
Orchestrators can be used with most deployment or virtualisation strategies such as Docker, bare metal deployments or cloud infrastructure.
cluster health, system metrics and business intelligence insights
Monitoring collects and aggregates system and process level metrics for each node within the cluster as well as detailed insights about connected users, subscribed topics, load distribution and resource allocation.
It offers endpoints and integrations for popular timeseries databases and monitoring dashboards such as Prometheus, InfluxDB, OpenTSDB, Cloudwatch, Datadog, Elasticsearch, Logstash, Kibana, Grafana, Graphite, Riemann, or Splunk.
monitoring overviewgateways allow clients to interact with deepstream ressources using a range of protocols
deepstream’s own distributed realtime protocol (patent pending) is optimized for speed and maximal flexibility at minimal throughput
as an ubiquitous IoT standard mqtt is optimized for low powered devices and unstable network conditions
http extends the range of deepstream compatible endpoints and allows for easy bulk processing of resource interactions
the websocket protocol is frequently used to interface deepstream endpoints with legacy systems
binary support enables endpoints to share files or media streams and allows for binary serialization with MsgPack or Protobuff
the query language developed by facebook makes reading and writing relational resources within deepstream easy
gateways allow clients to interact with deepstream ressources using a range of protocols
We know realtime – so much is fair to say. But that also means that we know how different it can be to run large scale realtime architectures in production. Whether it’s help with infrastructure, architecture, deployment, HA/failover setups, load balancing, routing or security – we’ll be happy to be on your side.
Using deepstream is straight forward, but creating large scale realtime architectures can be a bit fiddly. Changes can come in at any time from any direction and logic needs to be executed in a myriad of possible states. Our engineers have years of experience building large scale realtime apps and are happy to provide online support for your developers at every step along the way.
We help you kickstart your project. Your team gains first hand expertise from deepstream’s (surprisingly nice and sociable) core developers and learns about the intricacies of the platform, the correlation of concepts, the most efficient development patterns and how to keep performance at maximum.
Building your app is what you're best at. We'll leave that to you. But if you need help, we’d be happy to provide bespoke integrations with your internal systems, identity providers, messaging systems, databases and caches or create other features that help you make the most out of deepstream within your environment.