Do you know that you can access more real exam questions via Premium Access? ()
An Architect needs to automate the daily Import of two files from an external stage into Snowflake. One file has Parquet-formatted data, the other has CSV-formatted data.
How should the data be joined and aggregated to produce a final result set?
Answer : B
According to the Snowflake documentation, tasks are objects that enable scheduling and execution of SQL statements or JavaScript user-defined functions (UDFs) in Snowflake. Tasks can be used to automate data loading, transformation, and maintenance operations. Snowflake scripting is a feature that allows writing procedural logic using SQL statements and JavaScript UDFs. Snowflake scripting can be used to create complex workflows and orchestrate tasks. Therefore, the best option to automate the daily import of two files from an external stage into Snowflake, join and aggregate the data, and produce a final result set is to create a task using Snowflake scripting that will import the files using the COPY INTO command, and then call a UDF to perform the join and aggregation logic. The UDF can return a table or a variant value as the final result set.Reference:
A company has a source system that provides JSON records for various loT operations. The JSON Is loading directly into a persistent table with a variant field. The data Is quickly growing to 100s of millions of records and performance to becoming an issue. There is a generic access pattern that Is used to filter on the create_date key within the variant field.
What can be done to improve performance?
Answer : A
The correct answer is A because it improves the performance of queries by reducing the amount of data scanned and processed. By adding a create_date field with a timestamp data type, Snowflake can automatically cluster the table based on this field and prune the micro-partitions that do not match the filter condition. This avoids the need to parse the JSON data and access the variant field for every record.
Option B is incorrect because it does not improve the performance of queries. By adding a create_date field with a varchar data type, Snowflake cannot automatically cluster the table based on this field and prune the micro-partitions that do not match the filter condition. This still requires parsing the JSON data and accessing the variant field for every record.
Option C is incorrect because it does not address the root cause of the performance issue. By validating the size of the warehouse being used, Snowflake can adjust the compute resources to match the data volume and parallelize the query execution. However, this does not reduce the amount of data scanned and processed, which is the main bottleneck for queries on JSON data.
Option D is incorrect because it adds unnecessary complexity and overhead to the data loading and querying process. By incorporating the use of multiple tables partitioned by date ranges, Snowflake can reduce the amount of data scanned and processed for queries that specify a date range. However, this requires creating and maintaining multiple tables, loading data into the appropriate table based on the date, and joining the tables for queries that span multiple date ranges.Reference:
An Architect has a design where files arrive every 10 minutes and are loaded into a primary database table using Snowpipe. A secondary database is refreshed every hour with the latest data from the primary database.
Based on this scenario, what Time Travel query options are available on the secondary database?
Answer : A
Snowflake's Time Travel feature allows users to query historical data within a defined retention period. In the given scenario, since the secondary database is refreshed every hour, Time Travel can be used to query each hourly version of the table as long as it falls within the retention window. This does not include individual Snowpipe loads within each hour unless they coincide with the hourly refresh.
An Architect for a multi-national transportation company has a system that is used to check the weather conditions along vehicle routes. The data is provided to drivers.
The weather information is delivered regularly by a third-party company and this information is generated as JSON structure. Then the data is loaded into Snowflake in a column with a VARIANT data type. This
table is directly queried to deliver the statistics to the drivers with minimum time lapse.
A single entry includes (but is not limited to):
- Weather condition; cloudy, sunny, rainy, etc.
- Degree
- Longitude and latitude
- Timeframe
- Location address
- Wind
The table holds more than 10 years' worth of data in order to deliver the statistics from different years and locations. The amount of data on the table increases every day.
The drivers report that they are not receiving the weather statistics for their locations in time.
What can the Architect do to deliver the statistics to the drivers faster?
Answer : B
To improve the performance of queries on semi-structured data, such as JSON stored in a VARIANT column, Snowflake's search optimization service can be utilized. By adding search optimization specifically for the longitude and latitude fields within the VARIANT column, the system can perform point lookups and substring queries more efficiently. This will allow for faster retrieval of weather statistics, which is critical for the drivers to receive timely updates.
A new table and streams are created with the following commands:
CREATE OR REPLACE TABLE LETTERS (ID INT, LETTER STRING) ;
CREATE OR REPLACE STREAM STREAM_1 ON TABLE LETTERS;
CREATE OR REPLACE STREAM STREAM_2 ON TABLE LETTERS APPEND_ONLY = TRUE;
The following operations are processed on the newly created table:
INSERT INTO LETTERS VALUES (1, 'A');
INSERT INTO LETTERS VALUES (2, 'B');
INSERT INTO LETTERS VALUES (3, 'C');
TRUNCATE TABLE LETTERS;
INSERT INTO LETTERS VALUES (4, 'D');
INSERT INTO LETTERS VALUES (5, 'E');
INSERT INTO LETTERS VALUES (6, 'F');
DELETE FROM LETTERS WHERE ID = 6;
What would be the output of the following SQL commands, in order?
SELECT COUNT (*) FROM STREAM_1;
SELECT COUNT (*) FROM STREAM_2;
Answer : C
In Snowflake, a stream records data manipulation language (DML) changes to its base table since the stream was created or last consumed. STREAM_1 will show all changes including the TRUNCATE operation, while STREAM_2, being APPEND_ONLY, will not show deletions like TRUNCATE. Therefore, STREAM_1 will count the three inserts, the TRUNCATE (counted as a single operation), and the subsequent two inserts before the delete, totaling 4. STREAM_2 will only count the three initial inserts and the two after the TRUNCATE, totaling 3, as it does not count the TRUNCATE or the delete operation.
