5. CONCEPTS
▪ Event-processing engine
▪ In memory cluster
▪ PaaS - Serverless
▪ SQL Like programming
language
▪ JavaScript per UDF/UDA
▪ Availability 99,9%
▪ Scale-Up/Down
6. INPUT
EventHub
EventProcessedUtcTime, EventEnqueuedUtcTime, PartitionId
IoT Hub
IoTHub.MessageId, IoTHub.CorrelationId…..
BlobStorage
Blob vengono presi in considerazione ogni secondo e solo una volta
Pattern cluster1/logs/{date}/{time}/{partition}
Reference Data
Pattern products/{date}/{time}/product-list.csv
7. QUERY LANGUAGE
SELECT
Make,
System.TimeStamp AS Time,
COUNT(*) AS [Count]
INTO
AlertOutput
FROM
Input TIMESTAMP BY Time
GROUP BY
Make,
TumblingWindow(second, 10)
HAVING
[Count] >= 3
▪ Aggregate Functions (SUM, AVG…)
▪ Analytic Functions (LAG, LAST, ISFIRST…)
▪ Array Functions
▪ Conversion Functions (TRY_CAST, GetType)
▪ Date Time Functions (DATEADD, DATEDIFF, DAY, MONTH…)
▪ GeoSpatial Functions (ST_DISTANCE, ST_WITHIN…)
▪ Mathematical Functions (ABS, SIGN…)
▪ Record Functions (GetRecordPropertyValue…)
▪ String Functions (LEN, SUBSTRING, CONCAT…)
8. OUTPUT
Data Lake Store
SQL Database (NO SQL Server su VM)
Blob Storage
Table Storage
Cosmos DB
ServiceBus Queue / Topic
Event Hub
Functions
Power BI
17. STREAM ANALYTICS ON IOT EDGE
▪ Low-latency command and
control
▪ Limited connectivity to the
cloud
▪ Limited bandwidth
▪ Compliance
18. STREAM ANALYTICS ON IOT EDGE
▪ ASA on IoT Edge offers a cross-platform
solution running on Docker containers
Linux or Windows
▪ Multiple architectures (Intel or ARM
CPUs)
▪ Raspberry Pi
▪ Industrial PCs
▪ Dedicated field gateways
▪ Servers
20. QUERY LANGUAGE
▪ Most SQL query language features are already supported.
▪ Following feature are NOT yet supported:
▪ User-defined functions
▪ Azure ML functions
▪ Using more than 14 aggregates in a single step
▪ AVRO format for input/output
▪ Geospatial operators
▪ PARTITION BY
▪ GetMetadataPropertyValue