I have created azure devops with a few organizations/projects in it. Do I need to scale or replicate it across different datacenters so all the developers across the globe can access it? When would I need to scale or replicate a service? Any references appreciated.
Azure DevOps is already replicated across multiple Azure Data Centers. They did this after the south east data center hosting Azure Devops (then VSTS) went down for a few days, bringing down the service. By going with a cloud provider, this is one of the benefits you (assume) you get. Automatic (geo)replication/redundancy of SAAS products. You don't have to manage this, but if something goes wrong, you have no control (or responsibility) to fix it.
Manual intervention is not required for your Azure Devops organisation to be accessible to different developers accross the globe.
The on-premises version of Azure Devops can be scaled with multi-server deployments. However it's common to only deploy these servers to a single datacenter.
Do I need to scale or replicate it across different datacenters so all the developers across the globe can access it?
- No, you do not need to scale or replicate your data or projects to be accessable globally; however, some countries may have laws in place that prevent access to your project that reside outside of their region.
- The notion of renting someone else's infrastructure is that typically the infrastructure provider offers some degree of replicating your organization's data. Meaning replication of your data generally isn't needed.
- Applications/projects having failures or going down can happen and Microsoft's Azure documentation has suggestions on managing backups and disaster recovery. They also have services on Site Recovery in general.
When would I need to scale or replicate a service?
- When you want to minimize the effects of any single service having down time or being in a none recoverable state.
- When Availability Zones becomes something needed in respects to the size of your projects. An example is needing to scale and replicate your organizations/projects across multiple data centers to ensure high availability is maintained.