We are considering going from an Unlimited license to Enterprise license. While the Unlimited licence comes with a full sandbox, it is a paid upgrade with the Enterprise license.

The most obviously impact is we won't have production data in our developer sandboxes.

Are there any development, deployment, testing, and/or admin functions we won't be able to do with a developer sandbox that we are able to do with a full sandbox?

I'm aware of the documents Salesforce provides describing the recommended uses and differences in limits of each sandbox type. What I'm looking for is the practical, real-world issues faced by developers and admins whose organization dropped their full sandbox and were left using only developer sandboxes.

  • 2
    A big loss is the ability to replicate issues in production within a sandbox. It all depends on your structure. Say a specific record is causing issues. The data that exists can be difficult to recreate exactly and the issue may or may not appear by manually creating the data in a developer sandbox. Debugging and modifying code on the fly is much easier in a full sandbox as it most closely represents what production looks like. Also, in some cases devs do not have access to production so debugging in a full sandbox is helpful. Otherwise most everything else can be done in a dev sandbox.
    – Eric
    Commented Apr 20, 2016 at 15:23

2 Answers 2


Here is my practical take on this based on real world experience:

Without a fullcopy sandbox, the following becomes much riskier/harder

  • testing the migration of data from schema A to schema B while schema A has to stay 'alive'. This can come up when you simply have to change your existing schema in some way while keeping the org running. Akin to changing an engine while the plane is still flying. A fullcopy sandbox allows you to verify that existing business processes still work while you test out new functionality. The fullcopy sandbox also allows you to verify that data conversions don't break when applied to all of the existing data
  • testing the performance of some batch process or anything where there is a long-running executable, including a query
  • as a backstop for testing where your regression suites aren't 100% robust
  • As a diagnostic and testing platform for something in PROD that isn't working right but where the normal PROD debugging tools are overwhelmed by the data volume and you need to inject more diagnostic code or experiments into a PROD replica.
  • A place to test the inbound migration/merger of some acquired subsidiary's SFDC org

Personally, if the higher ups in my org decided to ditch the fullcopy sandbox, I'd turn to opiate abuse. A fullcopy sandbox has saved my bacon more than once.


As described in the Sandbox Types documentation:

Different sandbox types support different activities.

Developer Sandbox

A Developer Sandbox environment isolates changes under active development until they’re ready to be shared. A Developer Sandbox is designed for a single developer, although more than one person can log in to the sandbox at a time.

Developer Pro Sandbox

A Developer Pro Sandbox environment provides the same functionality as Developer Sandboxes, but offer increased file and data storage. A Developer Pro Sandbox can host larger data sets than a Developer Sandbox. Use a Developer Pro Sandbox for tasks such as data loading and integration testing, user training, and the development tasks you do in a Developer Sandbox.

Partial Copy Sandbox

Partial Copy Sandbox environments include all your org’s metadata, as well as a sample of your production org’s data that you define by using a sandbox template. To create a Partial Copy Sandbox, apply a sandbox template at creation time. Use Partial Copy Sandboxes for virtually any development, testing, or training purpose. The only tasks for which they aren’t well-suited are full performance and load testing.

Full Sandbox

A Full Sandbox is an environment with a replica of your entire production org and all its data. Use Full Sandboxes for any development, testing, or training purpose, including full performance and load testing. Apply a sandbox template so that your sandbox contains only the records that you need for testing or other tasks.

To sum it up, the major difference between each sandbox is the refresh interval, the available resources and the type of data that is being copied. If you are willing to provide training to your employees or perform more specific tests such as integration testing then having real data could be very beneficial for your business cases. So depending on your company size and the use cases you'd want to perform each Sandbox has it's own advantages.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .