Azure Time Series Insights pricing

Customers have provided feedback that they would prefer pricing that allows them to scale data processing, storage and querying independent of each other. As we introduce the revamped Time Series Insights service with the new set of industrial IoT insights capabilities, we are introducing a new business model that aligns pricing with our customer usage patterns.

Microsoft will continue to support customers who are using the S1 and S2 SKUs. However, we encourage customers move to the new business model as it offers several benefits – new set of industrial IoT insights capabilities that come with lower price points, lower total cost of solution, flexibility to pay-per-use and ability to control costs by changing data storage and querying patterns.

Consider the scenario where a customer has 10,000 devices each of which generate 5 MB of data per month.

The data processed is 10,000*[5 MB]/[1,024 MB/GB] = 48.83 GB. Since the Data Processing unit allows you to process up to 100 GB of data, you pay only $-/month for data processing.

Of the data that is processed, if you configure 10 GB for warm data analysis and analysis of data requires 1 query per month over the 10 GB of data stored, then

Warm Storage charges = 10 GB*$-/GB = $- per month

Warm Queries = No charge

If the remaining data is sent to cold storage and is queried once a month as well, then

Cold Storage charges = 38.83 GB*$-/GB = $- (Uses Blob pricing)

Cold Queries = 38.83 GB*$-/GB = $- per month

Total Monthly Charges for Time Series Insights = $- for Data Processing + $- for Warm Storage + $0 for Warm Queries + $- for Cold Storage + $- for Cold Queries = $-.

The new SKU introduces the new timeseries-optimized cold path analytics while continuing to support warm path ad-hoc data exploration capabilities. The service, at the time of public preview launch will start by supporting the cold path analytics scenarios. The service will be updated in the coming months to fold in the in-market warm path, ad-hoc analytics with the ability to route timeseries data to either one or both storage layers as your business requires.

The user experience for setting up your time series environment will include the current S1/S2 SKUs as well as the updated new SKU for the time being. We expect that over time, as the warm adhoc analytics capabilities get folded into the new SKU, we will encourage customers to move away from S1/S2 SKUs into the single, unified SKU and capabilities in the Time Series Insights offering to leverage the updated pricing benefits as well as more comprehensive capabilities of the end-to-end solution.

Each unit of S1 includes a maximum of 30 GB of total storage or 30 million events (whichever comes first), as well as maximum daily ingress and configurable retention. Also included is the Time Series Insights visualization experience. There is no limit on the number of users or queries that can be conducted.

For example, If you provision 1 unit of S1, you can configure your environment to retain data for up to 400 days of retention. If configured to 400 days retention, you cannot exceed an average daily ingress of 1/400th of the allotted environment (for 1 unit of S1, 30 GB). In this example, 1/400th of the total storage capacity per day is 0.075 GB or 75 MB.

Another example—if you configure your environment to retain data for 180 days with one S1 unit, you cannot exceed daily ingress of 0.166 GB (166 MB) per day. If your ingress rate exceeds 0.166 GB per day, your data will not be stored for 180 days. If, for example, your ingress rate averaged 0.189 GB per day, you would realize just over 158 days of retention (30 GB/0.189 = 158.73 days of retention).

If you want to maximize your daily ingress, you can ingress up to 1 GB per day, and retain that data for up to 30 days.

Each unit of S2 includes a maximum of 300 GB or 300 million events (whichever comes first) of total storage and configurable daily ingress and retention. Also included is the Time Series Insights visualization experience. There is no limit on the number of users or queries that can be conducted.

For example, if you provision 1 unit of S2, you will achieve 400 days of retention if you configure the environment to retain for 400 days and your average daily ingress is 1/400th of 300 GB (0.75 GB or 750 MB per day) or less.

Another example—if you configure your environment to retain data for 180 days with one S2 unit, you will achieve 180 days of retention so long as you do not exceed daily ingress of 1.66 GB (1,660 MB) per day. If your ingress rate exceeds 1.66 GB per day, your data will not be stored for 180 days. If for example, your ingress rate averaged 1.89 GB per day, you would achieve just over 158 days of retention (300 GB/1.89 = 158.73 days of retention).

If you want to maximize your daily ingress, you can ingress up to 10 GB per day, and retain that data for up to 30 days.

An event is a unit of data up to 1 KB. Larger events are counted in multiples of 1 KB. For example, 0.8 KB is billed as one event, but a 2.6 KB event is billed as three events. Events larger than our maximum, 32 KB, will be rejected.

You can change the number of units via the Azure Portal.

You will be billed on a daily prorated basis. A day commences at midnight UTC for the purposes of billing. The billing is generated on a monthly basis. For instance, if a customer uses three units of Time Series Insights S1 for a total of 5 days, the customer is billed at the end of the month for 3 units × 5 days × ($-/31) = $-.

Similarly, if a customer uses three units of Time Series Insights S2 for a total of 5 days, the customer is billed at the end of the month for three units × 5 days × ($-/31) = $-.