Running several tenants
What this section is for
One Salesforce org can talk to several FrontSpin accounts. This section is about doing that safely — and in particular about not switching on enforcement until the configuration is ready for it.
If you have one FrontSpin account, you can skip the whole section. The default mode is exactly right for you.
The three modes
The org runs in one of three modes. The mode decides how strictly per-Record-Type routing is applied.
| Mode | Routing rules are | Traffic uses | Unmapped records are |
|---|---|---|---|
| LEGACY | Ignored for traffic | The original single tenant | Sent as before |
| CONFIGURATION | Loaded and checked | The original single tenant | Sent as before |
| STRICT | Enforced | The tenant the rule names | Refused, not guessed |
LEGACY is the default, and it is the fail-safe. The org only reaches STRICT by somebody deliberately setting it.
Why CONFIGURATION exists
It is the rehearsal. In CONFIGURATION mode you can load every routing rule, run the health check, and see exactly what STRICT would do — while traffic carries on going where it always went and nothing is refused.
Going straight from LEGACY to STRICT without that rehearsal is how an org discovers at nine on a Monday that fourteen Record Types have no rule.
How the mode is set
Setup → Custom Metadata Types → FrontSpin Setting → Manage Records
There should be exactly one active row, with Routing Mode set to LEGACY,
CONFIGURATION or STRICT.
Everything ambiguous means LEGACY
This is worth knowing in full, because it means a configuration mistake can never accidentally switch enforcement on:
| Situation | Resulting mode |
|---|---|
| No setting row at all | LEGACY |
| The row is inactive | LEGACY |
| Two or more active rows | LEGACY |
| Routing Mode is blank | LEGACY |
| Routing Mode is a word nobody recognises | LEGACY |
So if you set STRICT and nothing changes, the first thing to check is whether a second active row exists. Two active rows is not an error — it is silently LEGACY.
Moving off LEGACY
- Create the instances and routing rules you need.
- Set the mode to CONFIGURATION.
- Run the health check. Fix every ERROR.
- Work through the WARNING lines. Each one is a Record Type that STRICT would refuse.
- When the check is clean, set the mode to STRICT.
- Watch sync errors closely for a day.
You can go back. Setting the mode to LEGACY restores the previous behaviour immediately, with no deploy. If STRICT causes trouble you cannot diagnose quickly, switch back and investigate calmly.
What goes wrong
| Symptom | Cause |
|---|---|
| STRICT was set, nothing changed | Two active setting rows, or a typo in the mode. |
| Records suddenly refused after a change | STRICT is on and a Record Type has no rule. |
| The health check is clean but traffic still goes to one tenant | Mode is CONFIGURATION, not STRICT. That is CONFIGURATION working correctly. |
Where this fits
Instances and routing rules are what STRICT enforces. The eight outcomes are what it reports.