4.3 out of 5 (21 Ratings)

20 Verified Reviews

Amazon Redshift

Easy to deploy, but AWS service slow

What worked well - using the Platform as a service - no need for indexing, partitioning etc - amazing compression - MAtillion (an AWS Marektplace ELT tool) proved very good at transferring SQL Server to Redshift What didn;t work well - poor support (AWS are too busy/successful) - performance not as stellar as had been led to expect. A significant level of specialist tuning is still required - even when top of range hardware deployed for demo-

Amazon Redshift

Very fast, but some of the database concepts were not implemented correctly


Amazon Redshift

Redshift delivers on promise of fast performance

We experienced very fast query performance. The ease of implementation was great.

Amazon Redshift

AWS RedShift Great Service Improvement

Great documentation and well articulated guides on implementation. RedShift fits well into existing AWS VPC investments and security practices.


AWS Challenges

It was difficult to speak directly with an AWS technical person. It was much easier through a 3rd party that had a relationship with AWS.

Amazon Redshift

An Excellent system for the price. Low start up cost.

Redshift is overall a good product. We are willing to live with its shortcomings for the overall lower cost. It has served us well so far.

Amazon Redshift

Great Implementation

Great - positive experienece

Amazon Redshift

Easy to get started, tricky to master in depth

Pro: well documented, built on standards (PostSQL), so easy to learn Cons: requires deep query expertise to ensure queries complete in a reasonable amount of time

Amazon Redshift

AWS Redshift

AWS has made our companies growth extremely easy. Redshift has proven to be a valuable resource for us.

Amazon Redshift

Easy to get started and scale, but only once you've RTFM

Easy to turn on, deploy, and experiment with. With AWS Redshift, an important piece of the technology is that it works best in a cluster. Single nodes aren't performant at all due to how the horizontal scaling architecture works, but are fine for development. Additionally, scaling writes has proved challeging. At launch, we appended data to 100 tables every 15 minutes. But, once read load was applied, it brought the cluster to a hault. Reducing our load intervals and increasing our cluster size solved all issues.