Paolo Di Tommaso
Paolo Di TommasoJun 02, 2020

Tower Cloud Launch

Today marks quite an update in the Nextflow Tower project. The months in lockdown have been intense for our team, and it is a credit to everyone that the development efforts have continued at full pace. Since its launch in September, Tower has monitored more than 55 million user jobs with almost zero downtime! We have been very excited by its adoption and would like to thank all the early testers and users.

Now, it’s time to move on. In the past few months, we have worked intensely on a series of new features and today we are delighted to be releasing some of these.

GitHub sign-in authentication

Since the launch of Tower, we didn’t want to have yet another annoying sign-up form
to access the Tower dashboard. We also wanted to avoid having yet another password.

Today we are making it even easier to sign up and log in. All users with a GitHub account (and we know our users love it!) will be able to authenticate Tower using GitHub single sign-on.

In the login form click Sign in with GitHub to access Tower (you don’t need to fill out the email field either).

If you have already used Tower in the past and you want to continue to access your past pipeline executions, make sure the GitHub account uses the same email address as the one you used to access Tower.

Tower launch!

Tower Launch has finally landed. This feature allows you to deploy the execution of Nextflow pipelines using a streamlined web interface into popular cloud services!

barcelona offices

When clicking the Launch button, a dialog form allows you to specify: which compute environment to deploy the execution, the repository URL of the pipeline to deploy, and a few more details such as the pipeline version, the Nextflow configuration profile(s), and the cloud bucket you want to use to for the working directory.

To provide the input data, this is achieved via the pipeline config input box or profile instead of uploading data files. This approach streamlines the deployment procedure and above all, makes the overall process easily replicable.

Today we are releasing with support for AWS Batch and Google Cloud Life Sciences compute services. More cloud and batch schedulers will be added shortly!

Re-launch and execution logs

Along with the launch, Tower also provides Re-launch which allows you to re-deploy the execution of previously executed pipelines. It spares having to re-enter the launch configuration and helps to make each run replicable tracking relevant information such as the project Git commit id and the pipeline work directory.

Moreover, the pipeline log can be instantly accessed during the pipeline execution if you choose a compute service providing the support for streaming such as AWS CloudWatch.

Private repos and security

Today’s release also includes support for private pipeline repositories hosted in popular services such as BitBucket, GitLab and GitHub. More providers will follow.

To access private repositories, you will need to provide the security credentials for such service.

At this regard, it’s important to highlight that Tower securely stores all secrets & credentials using a 256-bit AES encryption algorithm.

Conclusion

We are very excited about today’s launch. It marks a significant step forward in our vision of simplifying data analysis pipelines. If your organisation is interested in deploying Tower in your cloud or on-prem environment, please reach out to us, and we would be happy to explore options.

A big thanks to all the early adopters, to those who provided feedback and in particular to the Nextflow community and the nf-core team for all their continued support.