Standard Octopus server installations run on a single central server node. When that node fails, or needs to go offline for Windows Updates, the Octopus service is unavailable.
With Data Center, Octopus instances can be load balanced and run across a cluster of nodes, ensuring the Octopus service is highly available. This allows Octopus server nodes to be temporarily disabled for operating system updates, and provides redundancy in the case of node failure.
Used alongside a clustered SQL Server database, this configuration allows you to keep the Octopus service available 24 hours a day, 7 days a week.
Deploy Octopus at scale
Having a consistent continuous delivery and DevOps story across your enterprise just makes sense. It makes it easy for people to change between teams, allows for greater knowledge sharing, allows you to balance the benefits of team autonomy and "self-service" with managing the risks around security and access to production.
With Octopus Data Center, all teams can use a single, highly available Octopus instance, or you can give each team their own private, isolated Octopus instance.
Web and API requests are load balanced between Octopus server nodes using any standard software or hardware load balancer.
When a deployment is executed, it is taken by a single Octopus server node from start to finish. As more deployments are queued, each Octopus node will take deployments from the queue. As long as the network and shared storage can keep up, the result is near-linear scalability in performance as nodes are added. Our performance testing found that a two-node Octopus cluster was able to execute 100 large deployments twice as quickly as a single-node cluster could.
Give teams their own space Q3, 2018
As more and more teams use Octopus, the number of projects and environments can appear overwhelming - especially if you prefer to give teams view-access to projects beyond their own.
"Spaces" will make it easy for team members to focus on only the projects that matter to them, by grouping projects, environments and other settings into isolated spaces within Octopus. This keeps all the teams on a common platform while making it easier to concentrate on what matters to them.
Remote Release Promotions Coming soon
Hundreds of customers use Octopus today in PCI-DSS, SOX and HIPPA compliant environments with the right configuration. Proving that you are compliant can be made much easier when the production Octopus is kept completely isolated from the dev/test Octopus server. With Remote Release Promotions, you will be able to deploy releases in one Octopus server, then promote them to an isolated Octopus server, and continue to deploy them there.
Release Promotions will work whether your Octopus servers are divided by an air gap and you need to use a USB thumbdrive to promote them, or if there's one-way network access via a DMZ.
We have 350 applications and deploy to 2,700 web servers. We have 6 delivery centers across the globe that are using Octopus. We're a 24 hour shop 7 days a week. The availability of Octopus is very key to us. It needs to be up. — Jim Szubryt
Senior Manager at Accenture
You might also like...
Exensibility & REST API
The Octopus UI is built over a REST API, so anything the UI does, you can do too
@OctopusDeploy Really liking the new GUI in @OctopusDeploy 4.0, both good looking and fast. Great work!— Johan Boström (@zarx) November, 21 2017
So today I deployed a release using @OctopusDeploy to production on an international flight @ 35,000 feet. It felt weird and amazing all at the same time— Ben Macpherson (@BenjiMac93) May 30, 2018