Overall Satisfaction with Matillion
I met with Matillion team at AWS Summit 2018 at Sydney, Australia. At that time we were using legacy ETL for our Redshift Data Warehouse.
We were trying to adopt Continuous Integration/Continuous Delivery framework for our ETL and found it to be challenging for our environment. It was time to look for something new and say goodbye to our bellowed legacy ETL. So we had a conversation with Matillion team and quickly decided to do a POC. We've evaluated couple of other tools in the past but nothing came close to Matillion.
First of all, setup was a breeze. The migration was quick and painless, the system is very easy to work with. Almost more importantly, the Matillion team is very friendly and always willing to help, their support is outstanding! Matillion has dozens of native connectors out of the box, so we had no problem with integrations at all. In fact, migration to Matillion allowed us to drop couple of third party legacy ETL components required for AWS infrastructure.
Python scripts component allows to use Boto3 and hundreds of other libraries, so practically there is no limit of where you can use Matillion.
We were finally able to build full scale CI/CD pipeline. Our setup uses combination of JIRA, Bitbucket and Bamboo build server and it works like a charm, thanks to Matillion'™s REST API.
- Works very well with Redshift and integrates with other AWS Services, such as S3, SNS or SQS for example
- Has scripting components like Python using Boto and any other libraries. Additional libraries need s to be installed on Matillion EC2 instance
- Plenty of data sources out of the box, the rest can be pulled via API
- Automatic validation of database objects and components
- Easy to install
- Excellent integration with CI/CD
- Minor: Changes to the ETL can only be reviewed in Matillion GUI rather than true source code diff, i.e. Bitbucket
- Direct savings from actual running cost
- Indirect savings as a result of increased productivity
Matillion works very well for us. It simplified our old ETL and allowed to move ETL/Datawarehouse development process onto CI/CD framework.
We are still running on t2.medium instance, it copes with the load well. We do not use HA features at this stage, so cannot comment on that.
Fast and easy way to bring all of company's data into Redshift data warehouse. We haven't come across anything which Matillion cannot handle yet. We even control Redshift workload management (WLM) parameters from Matillion, that is we give ETL more Redshift resources during the actual ETL running time earlier morning and then take them back and give to analysts.