ENGAGEMENT
Right Start for Snowflake Optimization
CLIENT
Right Start
Engagement Summaries
Business Problem
Pixability has been using Snowflake for about 5 years and knew they were not using it as efficiently as possible. A painful example: Looker dashboards that took more than 25 minutes to populate. And increasing data volumes were followed by increasing costs.
Project Overview
Pixability engaged Right Triangle to take a detailed look at how they were using Snowflake and create a prioritized list of recommendations for improving everything from data ingestion to data transformation (primarily done within Snowflake) to data analysis (primarily done with Looker).
The goals of the project were to:
- Identify areas to improve Snowflake performance
- Identify areas to lower Snowflake costs
- Bring Pixability into alignment with Snowflake best practices
Performance Impact
We took 5 of Pixability’s longest running queries from the prior month and dramatically reduced the execution time of all of them:
| Execution Time | |||
|---|---|---|---|
| Query | Before | After | Time Savings |
| 1 | 1:14:57 | 1:12 | 98% |
| 2 | 47:08 | 1:54 | 96% |
| 3 | 29:43 | 1:55 | 94% |
| 4 | 24:51 | 1:48 | 93% |
| 5 | 25:23 | :20 | 99% |
| Average: | 96% | ||
Cost Impact
We compounded the performance impact by achieving an overall reduction in cost:
| Cost | |||
|---|---|---|---|
| Query | Before | After | Cost Savings |
| 1 | $11.99 | $3.07 | 74% |
| 2 | $7.54 | $4.86 | 36% |
| 3 | $4.75 | $4.91 | -3% |
| 4 | $3.98 | $4.61 | -16% |
| 5 | $4.06 | $0.85 | 79% |
| Average: | 43% | ||