Introduction to Snowflake (2023)

Cloud computing is now in its second decade of existence, and those two decades of development have been incredible. In these two decades, we have gone far beyond virtual machines and cloud storage. We now have tools like cloud functions (lambdas), containerization, and too many technologies to discuss in one article.

One area where there has been a lot of innovation is cloud databases. We have legacy database services that host familiar databases like MySQL, Postgres, and SQL Server. There are also offers for many other styles of data, such as documents, column data, key value data, and so on.

One of the relatively new cloud computing providers and the focus of this article is called Snowflake. Snowflake is a unique offering as it offers many features that developers need today. Do you need a cloud-based SQL database? Do you need a database that can query JSON data stored in a column? Do you need the ability to securely share data with external providers without exposing your infrastructure to them? Snowflake handles many of these concepts very well.

In my 30+ years of software development, I have seen many products come and go. Every once in a while I come across a product that I consider to be a game changer. The product I'm about to talk about is calledSnowflakeand I can say without a doubt that this is an innovative product.

Imagine a request from the marketing department. “We are going to be getting a large data set from Widgetemo, Inc. We will be conducting reviews on this and speedy processing is of paramount importance. This dataset will have around 100,000,000+ records and will be moved to Amazon's regular repository. I know this is not hard to use, but there is an exception. Isn't there always? This record contains PII (personally identifiable information) and can only be seen by the review team. Oh, and there's one more little item. We need to share the raw results with the client in real time every time we process the data. Please contact us with a quote.”

There are two ways to process this request. The first way is a classic cloud process. They add resources to EC2 (server) and RDS (database) instances to handle the load.

1. Create a process to download data from S3 into a set of objects.
2. Send this data to Postgres (SQL Server, MySQL, etc.).
3. Create a process to query the masked or unmasked data.
4. Create a process to send data to the customer's preferred sharing technology (downloadable file, SFTP, S3 bucket, Azure Storage, etc.).
5. Test the process and put it into production.

The second way to handle the load is to use some commands in Snowflake like:

1. add resources

ALTER LAGER SET WAREHOUSE_SIZE=GRANDE;

2. Import data from S3.

COPY TO S3 <COMMAND OPTIONS HERE>

3. Control personally identifiable information.

CREATE OR REPLACE MASK POLICY address_mask AS (string VAL) RETURNS string -> CASE WHEN current_role() IN ('MARKETING') THEN VAL ELSE '*********' END

4. Share data with the client.

CREATE SHARE <COMMAND OPTIONS HERE>

As you can see, each of these processes can fulfill the user's request. The big difference is the time from inquiry to production. The first process can take a day or more, depending on the backlog of the development team. The second process consists of a few commands from start to finish. Speed ​​is a clear competitive advantage, and it's Snowflake's built-in features that make it possible. We'll see.

snowflake features

Snowflake is a strange amalgamation of many of today's divergent database concepts. The following list represents some of the high-level features:

• It supports SQL and offers SELECT, INSERT, UPDATE, DELETE, etc., as well as CREATE TABLE, CREATE VIEW, CREATE SCHEMA and JSON queries.
• Snowflake can store and query JSON data stored in a special column type.
• It is cloud independent (Azure, AWS, GCP). I call it BYOC (Bring Your Own Cloud). You can set up your Snowflake infrastructure with the cloud provider of your choice.
• integrated python. Snowflake can embed and call Python code from your queries and procedures.
• Secure data exchange. Snowflake may share and use shared data with other instances of Snowflake. Computing costs in shared environments are paid by the consumer of that data, not the host.
• Access to multiple clients. You can access your data using the technology of your choice. There are drivers for ODBC, Python, Node, and GO.
• Pay as you go with "infinite" scaling. With the touch of a button (actually a simple command line), you can increase or decrease the performance of your computer for the queries made.

“Infinite” compute with Snowflake Warehouses

When working with Snowflake, you need to understand that the concept of a warehouse is not what you might think of as organized data storage. Instead, the concept of data storage in Snowflake is COMPUTE, not storage. Let me say it again: it is COMPUTING and not a data store.

A good metaphor is that it's like a car that you can add cylinders to as you go. Want to save gas ($$$)? Choose a small car. Speeding on the highway at 200 km/h? Convert the car to eight cylinders. Snowflake COMPUTE runs on powers of 2: 1, 2, 4, 8, 16, 32 and up to 256 cylinders of COMPUTE power. The unique aspect of Snowflake is that you can control how many cylinders you use for any given transaction. You pay for it, but the performance of updating a single query or set of queries with a switch is attractive.

Pay with the simple Snowflake model to pay credits based on the Snowflake edition you create. Credits are used to pay for credit hours: one credit = one credit hour. Based on the size of the camp, the system determines how credit hours are billed, and you pay more credits for larger camps. All queries are billed by the minute.illustration 1displays prices per credit hour per spin.

Illustration 1:The price you pay for Snowflake depends on the run and the size of your inventory.

Some features, such as PCI or HIPAA, require higher editions of Snowflake.

See Also

COPY TO| Snowflake-Documentation

To better understand how this works, let's look at an example. The following code is simple. Create new tables with different sizes of bearings. The source table is calledDEMO_DATAand contains 100,000,000 records. This code creates new copies of this table with different store sizes. The command is the following:

ALTERAR ALMACÉN SET WAREHOUSE_SIZE=PEQUEÑO; CREAR TABLA DEMO_DATA_SMALL COMO SELECCIONAR * DE PRARIE_DEVCON_DATABASE .PUBLIC.PEOPLE_HUGE; ALTER WAREHOUSE SET WAREHOUSE_SIZE=MEDIUM; CREATE TABLE DEMO_DATA_MEDIUM AS SELECT * FROM PRARIE_DEVCON_DATABASE .PUBLIC.PEOPLE_HUGE; ALTER WAREHOUSE SET WAREHOUSE_SIZE=LARGE; CREATE TABLE DEMO_DATA_LARGE AS SELECT * FROM PRARIE_DEVCON_DATABASE .PUBLIC.PEOPLE_HUGE; ALTER WAREHOUSE SET WAREHOUSE_SIZE=XXLARGE;CRIA TABELA DEMO_DATA_XXLARGE AS SELECT * FROM PRARIE_DEVCON_DATABASE .PUBLIC.PEOPLE_HUGE;

The results are the following:

• SMALL: 24 seconds
• MEDIUM: 19 seconds
• BIG: 12 seconds
• XXL: 8 seconds

As you can see, performance is important for larger bearing sizes. In some, the tables contain more than a billion records. The difference is incredible.

snow view

The main mechanism for managing your Snowflake infrastructure is a web-based application called Snowsight, which offers many features.Figure 2,Figure 3, miFigure 4Highlighting some of Snowsight's most used features.

Figure 2:The data canvas is used to create data elements such as databases, schemas, tables, and views.

Figure 3:The activity log is used to monitor the performance, success, and failure of commands run on Snowflake.

Figure 4:The administration screen is used to add users, groups, stores, etc.

understand spreadsheets

You can run scripts on the worksheet canvas.Figure 5displays a spreadsheet with a code snippet. This snippet is used to create the database, table, and user account that you'll use in the examples in this article.

CREATE DATABASE CODE_MAGAZINE_DEMO; create or replace TABLE CODE_MAGAZINE_DEMO.PUBLIC.DEMO_DATA ( UNIQUEID VARCHAR(200), LASTNAME VARCHAR(200), FIRSTNAME VARCHAR(200), ADDRESS VARCHAR(200), CITY VARCHAR(200), STATEORPROVINCE VARCHAR(200)) ;CREATE USER USER_MAGAZINE_CODE WITH PASSWORD='PASSWORD';GRANGE ROLE OF ACCOUNT ADMINISTRATOR FOR USER CODE_REVISTA_USUARIO

Figure 5:Snowsight's spreadsheet view is used to create, edit, run, and analyze Snowflake queries.

Create a data load pipeline

Before you can start querying, analyzing, and sharing data, you must transfer that data to your Snowflake instance. The next section shows how to create a simple pipeline so that you can access the data through Snowflake. Building a pipeline consists of the following steps:

1. Create a CSV file.
2. Send CSV to an S3 bucket.
3. Create a format in code.
4. Create a stage in code.
5. Create a COPY TO command.
6. Open a connection.
7. Run a SQL command to load data.

start your project

This example creates a small console application to get some sample data and send it to Snowflake. The first step is to create a console application using the Visual Studio interface. The following code examples are based on C# 11 and its new string functions. To enable this in your project, you need to add the following section to the top of your Visual Studio project file.

<LangVersion>latest</LangVersion>

After creating your project and enabling C#11, you can continue working onPipeline.add the codelisting 1to your application.listing 1is a class capable of creating a series of display objects. These sample objects are returned in aList<T>Structure that you can save to a CSV file that ultimately goes into the Snowflake database via an S3 bucket.

Listing 1 Code to create sample data

SnowflakeDemo namespace; public class RandomDataService { public class Person { public string UniqueId { get; Phrase; } = System.Guid.NewGuid().ToString(); public string lastname { get; Phrase; } = ""; public string name { get; Phrase; } = ""; public string address { get; Phrase; } = ""; public string city { get; Phrase; } = ""; public string StateOrProvince { get; Phrase; } = ""; } public List<Person> GetSamplePeople(int how many = 100000) { var lastNames = GetLastNames(); var firstNames = GetFirstNames(); var addresses = GetStreetNames(); var Cities = GetCities(); var states = GetStatesAndProvinces(); var retval = new List<Person>(); for (int i = 0; i < how many; i++) { var person = new Person(); Person.LastName = LastName[Between0and9()]; Person.Name = Names[Between0and9()]; Person.City = Cities[Between0and9()]; person.StateORProvince = states[Between0y9()]; person.address = $"{RandomAddressNumber()} " + $"{addresses[Between 0 and 9()]}"; retval.Add(Person); } return value; } public int RandomAddressNumber () { var random = Random.Shared.Next(1000, 99999); return randomly; } public int Between0and9() { var random = Random.Shared.Next(0, 9); return randomly; } public List<string> GetLastNames() { var retval = new List<string>(); retval.Add("Lucas"); retval.Add("Smith"); retval.Add("Spielberg"); retval.Add("Gygax"); retval.Add("garland"); retval.Add("Wolff"); retval.Add("West"); retval.Add("Kardashian"); retval.Add("Van Halen"); retval.Add("Grohl"); Refund Policy; } public List<string> GetFirstNames() { var retval = new List<string>(); retval.Add("Mary"); retval.Add("Leslie"); retval.Add("Jane"); retval.Add("Jessica"); retval.Add("John"); retval.Add("Pablo"); retval.Add("George"); retval.Add("Ringo"); retval.Add("Eddie"); retval.Add("Alex"); Refund Policy; } public List<string> GetStreetNames() { var retval = new List<string>(); retval.Add("Orange"); retval.Add("Main"); retval.Add("Maple"); retval.Add("oak"); retval.Add("poplar"); retval.Add("brown"); retval.Add("Elm"); retval.Add("Redwood"); retval.Add("Lincoln Boulevard"); retval.Add("Sepúlveda Boulevard"); Refund Policy; } public List<string> GetCities() { var retval = new List<string>(); retval.Add("Seattle"); retval.Add("Austin"); retval.Add("Regina"); retval.Add("Calgary"); retval.Add("Winnipeg"); retval.Add("Portland"); retval.Add("Los Angeles"); retval.Add("Oak"); retval.Add("Montreal"); retval.Add("Ottawa"); Refund Policy; } public List<string> GetStatesAndProvinces() { var retval = new List<string>(); retval.Add("AB"); retval.Add("SK"); retval.Add("CA"); retval.Add("O"); retval.Add("WA"); retval.Add("TX"); retval.Add("CO"); retval.Add("NY"); retval.Add("MN"); retval.Add("KY"); Refund Policy; }}

Send to CSV file

After you create your sample data, you can save these records to a CSV file. This example uses the CsvHelper library, which you can install using NuGet with the following command:

CsvHelper Installation Package -Version 30.0.1

After installing thePackage,add the codelisting 2. This class has a function that can write data to a CSV file with a specific delimiter. I prefer the vertical bar ("|") because commas are more common in the data than vertical bars.

Listing 2 Use CsvHelper to create a delimited CSV file

Using System.Globalization;Using CsvHelper.Configuration;Namespace SnowflakeDemo;public class CsvTools{ public static void WriteCsvFile(dynamic dataToWrite, string outputFile, string delimiter) { var config = new CsvConfiguration(CultureInfo.InvariantCulture); // include header config.HasHeaderRecord = true; // change the setting of the delimiter. Delimiter = delimiter; //configuración del separator de comillas.DeberíaCitar = args => true; using (var writer = new StreamWriter(outputFile)) using (var csv = new CsvWriter(recorder, configuration)) { csv.WriteRecords(dataToWrite); } }}

Send files to S3

After you create your CSV file, you must upload it to an Amazon S3 bucket. This is done using the AWS SDK S3 package, which can be installed via NuGet with the following command:

Installation package AWSSDK.S3 -Version 3.7.101.58

After installing the package, add the code fromlisting 3. This code uses the Amazon SDK to write your file to an S3 bucket. You need a few items: the bucket you want to write to, AWS Access, and the secret keys that grant access to that bucket.

Listing 3: Use the Amazon SDK to write the file to an S3 bucket

usando Amazon;Usando Amazon.Runtime;Usando Amazon.S3;Usando Amazon.S3.Transfer;Namespace SnowflakeDemo;public class AmazonTools{public AmazonS3Client GetS3Client(BasicAWSCredentials creds, RegionEndpoint endpointRegion = null) { var clientRegion = RegionEndpoint.USEast1 ; if (endpointRegion != null) { clientRegion = endpointRegion; } var cliente = new AmazonS3Client(creds, clientRegion); cliente de retorno; } public BasicAWSCredentials GetBasicAwsCredentials( string awsAccessKey, string awsSecret) { var retval = new BasicAWSCredentials( awsAccessKey, awsSecret); política de reembolso; } public void UploadFile(cliente AmazonS3Client, string BucketName, string filename) { var ms = FileToMemoryStream(filename); var Utility = new TransferUtility (Cliente); var req = new TransferUtilityUploadRequest(); var fi = nueva información de archivo (nombre de archivo); Utility.Upload(ms, NombreDeCubo, fi.Nombre); } public MemoryStream FileToMemoryStream(string fileName) { var fs = new FileStream(fileName, FileMode.Open, FileAccess.Read); var ms = nuevo flujo de memoria (); fs.CopyTo(ms); retorno ms; }}

Sending data to Snowflake

The final step in this process is to copy data from S3 to your Snowflake databases. This is done using Snowflake's COPY INTO command. The COPY INTO command has two sections that are necessary to make it easier to load your command. These are the STAGE and FORMAT sections of the command. We'll see.

A STAGE is a set of codes that provide the information needed to access your S3 data. Specifically, you need the bucket, item key, AWS access key, and AWS secret key. The following block of code represents a STAGE used to access data through an Amazon S3 bucket:

's3://[[BUCKET_NAME]]/[[ITEM_KEY]]' CREDENCIAIS = (AWS_KEY_ID='[[AWS_ACCESS_KEY]]' AWS_SECRET_KEY='[[AWS_SECRET_KEY]]') FORCE=true

Snowflake supports several types of data formats that can be used to load data into Snowflake tables. Snowflake supports CSV, JSON, XML, Parquet, and more. FORMATS describe how the data to be loaded into the file to be imported is displayed. The following code represents the format of a CSV file that is delimited by vertical bars and encloses the data in double quotes.

file_format=(type='csv' COMPRESSION='AUTO' FIELD_DELIMITER='|' RECORD_DELIMITER = '\n' FIELD_OPTIONALLY_ENCLOSED_BY= '"' SKIP_HEADER = 1 TRIM_SPACE = FALSE ERROR_ON_COLUMN_COUNT_MISMATCH = TRUE ESCAPE = 'NONE =' DATE_FORMAT = 'AUTO' TIMESTAMP_FORMAT 'AUTO' NULL_IF = ('\\N','NULL') )

The following code creates a STAGE fragment.

espacio de nombres SnowflakeDemo;public class SnowflakeTools{ public string GetStage( String BucketName, String ItemKey, String AccessKey, String SecretKey) { var retval = $""" 's3://{bucketName}/{itemKey}' CREDENCIALES = (AWS_KEY_ID=' {accessKey}' AWS_SECRET_KEY='{secretKey}') FORCE=true """; política de reembolso; }}

The following code creates a FORMAT snippet.

public string GetFormat(string delimiter){ var retval = $""" file_format=(type='csv' COMPRESSION='AUTO' FIELD_DELIMITER='{delimiter}' RECORD_DELIMITER = '\n' SKIP_HEADER = 1 TRIM_SPACE = FALSE ERROR_ON_COLUMN_COUNT_MISMATCH = TRUE ESCAPE = 'NONE' DATE_FORMAT = 'AUTO' TIMESTAMP_FORMAT = 'AUTO' NULL_IF = ('\\N','NULL') ) """; Retval de Rückkehr;}

Create a COPY TO command

Once you have STAGE and FORMAT, you can create your COPY INTO command. The following code can be used to facilitate this process.

public string GetCopyCommand( string databaseName, string tableName, string stageInfo, string formatInfo){var retval = $""" COPIAR EN {databaseName}.PUBLIC.{tableName} FROM {stageInfo} {formatInfo}; """; Retval de Rückkehr;}

The following code shows the last COPY INTO command that you can run from a database connection in a Snowsight worksheet.

COPIAR PARA CODE_MAGAZINE_DEMO.PUBLIC.DEMO_DATA FROM 's3://dashpoint-demo/SampleData.csv' CREDENCIALES = (AWS_KEY_ID='ACCESS_KEY'AWS_SECRET_KEY='SECRET_KEY') FORCE=true file_format=(type='csv' COMPRESSION='AUTO ' FIELD_DELIMITER='|' RECORD_DELIMITER = '\n' SKIP_HEADER = 1 TRIM_SPACE = FALSE ERROR_ON_COLUMN_COUNT_MISMATCH = TRUE ESCAPE = 'NONE' DATE_FORMAT = 'AUTO' TIMESTAMP_FORMAT = 'AUTO' NULL_IF = ('\\N','NULL') ) ;

Running ADO.NET Commands

After you create your COPY INTO command, you must call it to load your data into Snowflake. You can do it from a spreadsheet or from code.Figure 6displays the executed code in a worksheet.

Figure 6:COPY INTO code is executed from a spreadsheet

If you want to run this command from within your applications, you need to add another piece of code. This snippet is the connection string you use to call Snowflake from ADO.NET. The following snippet contains a method you can use to create a proper connection string:

string pública GetConnectionString(string snowflakeIdentifier, string userName, string password, string databaseName, string tableName) { var retval = $""" account={snowflakeIdentifier}; user={userName}; password={password}; db={databaseName} ;Esquema=öffentlich;Armazém=COMPUTE_WH """; Retval de Rückkehr;}

The following code must be known. It just opens a connection, creates a command object, and invokes it.ExecuteNonQuery(). I mentioned this before. Snowflake meets him where he is and admits the tools he's familiar with.

usando (var conn = new SnowflakeDbConnection() { ConnectionString = connStringCode } { conn.Open(); var cmd = conn.CreateCommand(); cmd.CommandText = copyCode; cmd.CommandType = CommandType.Text cmd.ExecuteNonQuery();

Der Pipeline-Code

You now have a nice set of pipeline code.listing 4displays the output of all the code you created earlier.Figure 7displays the data that you created in the pipeline process.

Listing 4: The complete pipeline code

using System.ComponentModel;Using System.ComponentModel.DataAnnotations;Using System.Data;Using CsvHelper;Using Snowflake.Data.Client;Namespace SnowflakeDemo{ inner class Program { static void Main(string[] args) { if (Environment .GetEnvironmentVariable( "DEMO_KEY") == null || Environment .GetEnvironmentVariable("DEMO_SECRET") == null || Environment .GetEnvironmentVariable("DEMO_SNOWFLAKE_USER") == null || Environment .GetEnvironmentVariable("DEMO_SNOWFLAKE_PASSWORD") == null) { console. WriteLine( "You need the following environment" + "Configuration variables:\n" + "DEMO_KEY\n"+ "DEMO_SECRET\n"+ "DEMO_SNOWFLAKE_USER\n"+ "DEMO_SNOWFLAKE_PASSWORD"); Console.ReadKey(); To give back; } var awsAccessKey = Environment.GetEnvironmentVariable("DEMO_KEY"); var awsSecret = Environment.GetEnvironmentVariable("DEMO_SECRET"); var snowflakeUser = Environment.GetEnvironmentVariable("DEMO_SNOWFLAKE_USER"); var snowflakePassword = Environment.GetEnvironmentVariable("DEMO_SNOWFLAKE_PASSWORD"); var snow cup identifier = "gra75419"; var nombreBaseDeDatos = "CÓDIGO_REVISTA_DEMO"; var table name = "DEMO_DATA"; var BucketName = "dashpoint-demo"; var fileName = @"D:\data\junk\CodeSampleData.csv"; var itemKey = "CodeSampleData.csv"; var random = new RandomDataService(); var csvTools = new CsvTools(); var amazonTools = new AmazonTools(); var SnowflakeTools = new SnowflakeTools(); var data = random.GetSamplePeople(250000); csvTools.WriteCsvFile(data,filename,"|"); amazonTools.UploadFile(amazonTools.GetS3Client( amazonTools.GetBasicAwsCredentials(awsAccessKey,awsSecret)), BucketName, filename); var step = snowflakeTools.GetStage(bucketName,itemKey,awsAccessKey, awsSecret); format var = snowflakeTools.GetFormat("|"); var copyCommand = snowflakeTools.GetCopyCommand(databaseName, tableName,stage,format); var connString = snowflakeTools.GetConnectionString( snowflakeIdentifier, snowflakeUser, snowflakePassword, databaseName, tableName); //Send data with (var conn = new SnowflakeDbConnection() { ConnectionString = connString }) { conn.Open(); var cmd = conn.CreateCommand(); cmd.CommandText = copyCommand; cmd.CommandType = CommandType.Text; cmd.ExecuteNonQuery(); } Console.WriteLine("Data sent to Snowflake"); Console.ReadKey(); } }}

Figure 7:An example of imported data using pipeline code

Protection of information through masking

One of Snowflake's most compelling features is its ability to control access to who can view the content of different data elements. This is especially important when it comes to PII information. This feature is known asmasking policyin snowflake The following command shows how to protect data using a masking policy. Create a masking policy that shows unmasked data for people with the MARKETING role and asterisks for everyone else. This policy is then applied to the ADDRESS column inDEMO_DATATisch.

CRIE OU SUBSTITUA POLÍTICA DE MASKING address_mask AS (string VAL) RETURNS string -> CASE WHEN current_role() IN ('MARKETING') THEN VAL ELSE '*********' ENDALTER TABLE DEMO_DATA MODIFY COLUMN ADDRESS SET MASKING POLICY dirección_mask;

Figure 8shows the unmasked data.

Figure 8:unmasked data

Figure 9displays the masked data that most users will see.

Figure 9:A sample of data imported using pipeline code in a masked state due to the selected ROLE via Snowsight

share data

Once you have data in Snowflake, you can share that data with other Snowflake users. The real benefit is that the consumer of that data pays for the computational costs. With other cloud databases, the data owner is responsible for usage costs.

When sharing data, you have several options. The first option is to share data with a consumer in the same region where your Snowflake instance is set up. This is by far the simplest mechanism for sharing data and the most common use case in my experience.

The second option is to share data with a consumer who lives in a different region or cloud provider. This type of sharing uses Snowflake replication tools. In this article, I'll explore the simple use case.

There are two basic types of actions. With output shares, you can share data from one or more Snowflake instances with other Snowflake users. Input Shares are data sources provided to you by other Snowflake users.

There are two ways to create an output share: using SQL code or using Snowsight. This article demonstrates the SQL code way to do this. The first is to run commands from a spreadsheet like this:

create share code_mag_outBound; Use the Code_Magazine_Demo database to share the use of Code_Mag_OutBoundGrant in the code_magazine_demo.public schema. ON SIGHT SELECT CODE_MAGAZINE_DEMO.PUBLIC.V_SORTED_DEMO_DATA TO SHARE CODE_MAG_OUTBOUND; OLD CODE SHARING_MAG_ADD OUTGOING ACCOUNTS=XXXXXX;

Once another Snowflake account has shared data with you, you will be able to access that data through a database created from the share. There are two ways to create a database from a share. The first is Snowsight. This is done by selectingShare private databy Snowsight. Then select "Shared with you" on the data sharing screen. This will open a screen with the records shared with you, as shown in theFigure 10.

Figure 10:The list of records shared with a Snowflake account

Click on the share for which you want to create a database. You will be presented with the Get Data dialog box as shown inFigure 11.

Figure 11:The Get Data screen is used to create external shared databases.

This dialog allows you to specify the name of your database. Optionally, you can also specify which security groups can access this database.

The second way to create a database from a share is to run a command from a Snowsight spreadsheet. To create a database from a share, run this command:

CREATE DATABASE OTRA_DEMO DAS PIEZAS XXXXX.YYYYY.CODE_MAG_OUTBOUND

The share name parameters for this command can be derived from the screen atFigure 11from the previous example.

After creating a database from a share, you can access it through Snowsight as shown inFigure 12.

Figure 12:The Snowsight Data screen showing the database created from a share

Figure 13displays shared data from demos created earlier in this article. One thing you'll notice right away is that the masking policy applies to shared data.

Figure 13:Masking policy applies to shared data

I've been working with Snowflake for several years and it wasn't until 2022 that I realized that the ability to seamlessly share data is monumental. We no longer need to set up S3 buckets or Azure blog storage accounts. We simply grant access to a set of tables/views and the client can access them using the SQL commands they already know.

Diploma

Snowflake is that rare commodity that has immense power but can be used by mere mortals to access that power. This article only scratches the surface of what Snowflake is capable of. Other areas of your interest may include Python integration, clean room technology, and other interesting use cases with another "snow" technology called Snowpipe. Look at her. You won't be disappointed.