Azure HDInsight offers fully managed and supported 100% Apache Hadoop®, Spark, HBase and Storm clusters. You can get up and running quickly on any of these workloads with a few clicks and within a few minutes without buying hardware or hiring specialized operations teams typically associated with big data infrastructure.
HDInsight clusters consist of a set of nodes. Customers will be billed for the usage of their nodes for the duration of the cluster’s life. Billing starts once a cluster is created and stops when the cluster is deleted and is pro-rated per minute. Customers can choose their cluster type and if they need Standard or Premium. If they need Premium, an additional surcharge will be added on top of the Standard price. Additionally, if customers choose R Server clusters, additional surcharges will also be reflected for these specific workloads. The preview prices below reflects a 50% discount on GA price for Premium and 75% discount on GA price for R-Server. Prices for R-Server will remain at Preview rates until January 1, 2017. HDInsight Kafka is also under limited Public Preview. For more information about the service and pricing, contact us.
The India Regions are available to customers with a business presence in India.
The India Regions are available to customers with a business presence in India. The West India data center is currently only available to select volume licensing customers and partners. For more information, contact your Microsoft India partner manager or account manager.
Azure Government is available to US government entities to purchase physically and network isolated instance of Azure Government from a Licensed Azure Government Service Provider or Partner with no upfront financial commitment or fee. Or, you can sign up for a free Azure Government trial.
Important: The price in R$ is merely a reference; this is an int'l transaction and the final price is subject to exchange rates and the inclusion of IOF taxes and an eNF will not be issued.
Azure Germany is available to customers and partners in the European Union (EU) and European Free Trade Association (EFTA) and provides data residency in Germany with additional levels of control and data protection with a modest price uplift over global cloud offerings (% varies per service).
HDInsight is not available in the Brazil South region. Please select another region.
HDInsight is not available in the South India region. Please select another region.
HDInsight is not available in the Canada East region. Please select another region.
HDInsight is not available in the Canada Central region. Please select another region.
HDInsight is not available in the UK West region. Please select another region.
HDInsight is not available in the UK South region. Please select another region.
HDInsight is not available in the West US 2 region. Please select another region.
HDInsight is not available in the West Central US region. Please select another region.
There is no surcharge in Germany for using R-Server until February 1, 2017. After February 1, 2017 regular GA rates will apply.
| HDInsight Feature | Standard Base Price /Node-Hour | Premium Base Price /Node-Hour + $0.02/Core-hour Preview | |
|---|---|---|---|
| Azure Cloud Data Platform | |||
| Unlimited Scale | |||
| Elastic scale | |||
| Easy provisioning | |||
| High availability | |||
| SLA | |||
| Enterprise Readiness | |||
| Basic Monitoring | |||
| Hadoop versions upgrades and patching | |||
| Encryption of data at rest | |||
| Secure Gateway and Zookeeper nodes | |||
| Ranger (secure Hadoop) + AD integration | |||
| Type of Clusters | |||
| Hadoop (Hive, Pig, Storm) | |||
| Hadoop with LLAP (interactive querying) | |||
| HBase for NoSQL Database | |||
| Storm for real-time stream analytics | |||
| Spark for in-memory SQL querying, ML, and real-time stream analytics | |||
| R Server for parallelized machine learning |
| Hadoop, HBase, Storm, Spark | R-Server | |
|---|---|---|
| Standard Cluster | Base Price /Node-Hour | Base Price /Node-Hour + $0.02/Core-hour |
| Premium Cluster Preview | Base Price /Node-Hour + $0.02/Core-hour | Base Price /Node-Hour + $0.04/Core-hour |
With HDInsight, full Hadoop support is included in your Azure support agreement. An Azure support subscription provides expert technical assistance from Microsoft for organizations implementing Hadoop and for all other services in Azure. It is also backed by Hortonworks, the leading contributor to Apache Hadoop open source. Hortonworks employs some of the original developers and architects of Hadoop who have contributed more than half of all code to Hadoop. This makes Microsoft more qualified to support your deployment through expert technical assistance and the ability to fix and commit code back to open source Hadoop.
Other Azure services associated with HDInsight such as Storage and Data Transfers are billed separately using standard rates. To estimate your bill, try the Azure Calculator.
Please note that HDInsight writes log files in the storage account associated with the cluster. These log files are available even after the cluster is dropped and can be deleted at any time.
The use of A1, A2, A3 nodes is limited to Zookeeper nodes and Headnodes. A-series is not available for data or worker nodes.
The use of A1, A2, A3 nodes is limited to Zookeeper nodes and Headnodes. A-series is not available for data or worker nodes.
| Instance | Cores | RAM | Disk Sizes | Base Price / Node |
|---|---|---|---|---|
| A1 | 1 | 1.75 GB | 70 GB | $0.08 |
| A2 | 2 | 3.5 GB | 135 GB | $0.16 |
| A3 | 4 | 7 GB | 285 GB | $0.32 |
| Instance | Cores | RAM | Disk Sizes | Base Price / Node |
|---|---|---|---|---|
| A4 | 8 | 14 GB | 605 GB | $0.64 |
| A5 | 2 | 14 GB | 135 GB | $0.35 |
| A6 | 4 | 28 GB | 285 GB | $0.71 |
| A7 | 8 | 56 GB | 605 GB | $1.41 |
Available in select data centers. A10 and A11 virtual machines feature Intel® Xeon® E5 processors. For high-performance clusters, modeling and simulations, video encoding, and other compute or network intensive scenarios. Similar to A8 and A9 instance configuration without the InfiniBand network and RDMA technology.
Note, when deploying a cluster on A10 or A11 machines, all nodes in the cluster must be A10 or A11
| Instance | Cores | RAM | Disk Sizes | Base Price / Node |
|---|---|---|---|---|
| A10 | 8 | 56 GB | 382 GB | $1.64 |
| A11 | 16 | 112 GB | 382 GB | $3.04 |
Dv2-series instances are the next generation of D-series instances that carry more powerful CPUs which are on average about 35% faster than D-series instances. It also carries the same memory and disk configurations as the D-series. Dv2-series instances are based on the latest generation 2.4 GHz Intel Xeon® E5-2673 v3 (Haswell) processor, and with Intel Turbo Boost Technology 2.0 can go to 3.2 GHz. Dv2-series. Customers running applications that require low latency, local SSD access, or require faster CPUs on Hadoop would benefit from using Dv2-series. HBase for HDInsight customers would benefit from the higher memory from Dv2-series to increase performance. HBase for HDInsight customers would benefit from the higher memory from D-series to increase performance. Storm and Spark for HDInsight customers would also benefit from higher memory for loading larger reference data and faster CPUs for higher throughput.
D1, D2, D11 nodes are only available as Zookeeper nodes for HBase and Storm. D-Series will continue to be available but the DV2-Series are recommended. Edge Node with R-Server is only available on D4 v2, D12 v2, D13 v2, D14 v2.
| Instance | Cores | RAM | Disk Sizes | Base Price / Node |
|---|---|---|---|---|
| D1 v2 | 1 | 3.5 GB | 50 GB | $0.155 |
| D2 v2 | 2 | 7 GB | 100 GB | $0.311 |
| D3 v2 | 4 | 14 GB | 200 GB | $0.622 |
| D4 v2 | 8 | 28 GB | 400 GB | $1.243 |
| D5 v2 | 16 | 56 GB | 800 GB | $2.486 |
| D11 v2 | 2 | 14 GB | 100 GB | $0.38 |
| D12 v2 | 4 | 28 GB | 200 GB | $0.76 |
| D13 v2 | 8 | 56 GB | 400 GB | $1.368 |
| D14 v2 | 16 | 112 GB | 800 GB | $2.462 |
HDInsight deploys different number of nodes for each cluster type. Within a given cluster type, there are different roles for the various nodes, which allow a customer to size those nodes in a given role appropriate to the details of their workload. For example, a Hadoop cluster can have its worker nodes provisioned with a large amount of memory if the type of analytics being performed are memory intensive.
Hadoop clusters for HDInsight are deployed with two roles:
HBase clusters for HDInsight are deployed with three roles:
Storm clusters for HDInsight are deployed with three roles:
Spark clusters for HDInsight are deployed with three roles:
The use of R-Server will incur one edge node in addition to the cluster deployment architecture.
We charge for the number of minutes your cluster is running, rounded to the nearest minute, not hour.
If you run a cluster for 100 hours in US East with 2 Head Nodes D13 v2, 3 Data Nodes D12 v2 and 3 Zookeepers D11, the billing would be the following in the two scenarios:
In order to stop an HDInsight cluster, you must delete the cluster. By default all data an HDInsight cluster operates on resides in Blob Storage, so data will not be impacted by this. If you want to preserve your Hive metadata (tables, schemas) you should provision a cluster with an external metadata store. You can find more details in this documentation
The number of data nodes will vary depending on your needs. With the elasticity available in Azure cloud services, you can try a variety of cluster sizes to determine your own optimal mix of performance and cost and only pay for what you use at any given time. Clusters can also be scaled on demand to grow and shrink to match the requirements of your workload.
Each subscription has a default limit on how many HDInsight data nodes can be created. If you need to create a larger HDInsight cluster or multiple HDInsight clusters that together exceed your current subscription maximum, you can request that your subscription's billing limits be increased. Please open a support ticket with Support Type = Billing. Depending on the maximum nodes per subscription that you request, you may be asked for additional information that will allow us to optimize your deployment(s).
To estimate the cost of clusters of various sizes, try the Azure Calculator.
