Cloud computing shapes how every Salesforce program works. Recognising the constraints early avoids surprises later.
- Design within governor limits from the start
Multi-tenancy means governor limits matter. Bulkify Apex, batch large data work, and respect SOQL and DML limits. Treat them as design constraints, not surprises to debug around.
- Plan for the three-times-yearly release cadence
Subscribe to Salesforce release notes. Run regression testing against the relevant sandbox release windows. Plan training and communications for major feature additions.
- Choose the right Salesforce region
For regulated workloads, pick a Salesforce region that meets your data-residency requirements. Hyperforce expanded regional choice; pick deliberately.
- Build integration on the API-first model
Every integration uses the same APIs the UI uses. Master REST, Bulk, and Pub/Sub for the bulk of integration work; reach for SOAP only when partner constraints require it.
- Use Trust.salesforce.com for operational visibility
Subscribe to the Trust portal feeds for the instance you run on. Real-time production status and planned maintenance live there; customer support tickets cite Trust data routinely.
- Multi-tenancy drives governor limits. Designing code without considering limits produces production failures.
- Three-times-yearly releases happen whether you are ready or not. Plan testing around the windows.
- Data residency requirements limit which Salesforce regions you can run on. Confirm before deployment.
- Cloud computing requires trusting the vendor. Salesforce publishes Trust data; use it for operational decisions.