Serverless Task — Snowflake — Part 2
Part1 edition describes about detailed about Serverless Task with Example?
let’s move on to the execution part of the study.
To compare their performance, I created a Server Task and a Serverless Task at the same time, and both tasks performed identical activities. I ran both tasks for approximately 18 hours.
Server task Output:
Serverless taks Output:
Output Comparison:
Query History: During the execution of the Serverless Task, the Snowflake-managed Warehouse used for execution is highlighted in a red rectangle box. The size of the Warehouse is X-Small.
Final Result : Credits consumed
One of the most important aspects of the comparison is the number of credits used by both the warehouses during the execution. To my surprise, the result shows a significant difference in credits used. Please find the result below:
Credits used by Server Task:
Credits used by Serverless Task:
Overall Comparison in credits:
Output: In this scenario, the Server Task has utilized 99% of the credits, while the Serverless Task has only utilized 1% of the credits, despite both tasks performing the same activity.
Workload Consumption of warehouse for server task (~18 Hours):
Another method of measuring the credit consumption of the Serverless Task.
Limitation of using this method : The data latency ranges from 45 minutes to 3 hours, depending on the view being used. However, the latency for the SERVERLESS_TASK_HISTORY view is 3 hours.
Recommendation (Shared by Snowflake Team)
I hope this article has provided you with some valuable insights into the Serverless Task feature. Please let me know if you require any further information.
References:-
About me:
I am a Data Engineer and Cloud Architect with experience as a Senior Consultant at EY GDS. Throughout my career, I have worked on numerous projects involving legacy data warehouses, big data implementations, cloud platforms, and migrations. If you require assistance with certification, data solutions, or implementations, please feel free to connect with me on LinkedIn.
