Skip to main content

Update and Repair Spotto AWS Access

Every connection method can be rerun safely. A rerun updates what's there and changes nothing when it's already current.

When You Need To Act​

Open the AWS account's edit page in Cloud Accounts. The Access row shows the deployed template version:

AccessMeaningDo this
Up to DateThe stacks run the latest templateNothing
Update AvailableA newer template adds permissions or fixesUpdate
Repair AccessThe role was changed by hand, lost policies or was deletedRepair

Missing-permission notices on an account link to the same guidance. You don't need to update for each Spotto release: AWS keeps ViewOnlyAccess and SecurityAudit current, so most new reads are already covered.

Update​

MethodHow
CloudShellCopy the command from Update Available and run it in the management account. It's the same command as the first connection: a rerun updates SpottoConnect-<deployment>, SpottoBillingAccess-<deployment> and SpottoMemberAccounts-<deployment> to the latest version and keeps every other setting.
AWS consoleCloudFormation (the access stack’s owning region) > SpottoConnect-<deployment> > Update > Replace existing template. Choose Amazon S3 URL, paste the versioned URL from Update Available, keep the parameters and submit. Repeat for SpottoMemberAccounts-<deployment>.
TerraformBump the module version, then terraform apply. Each module version pins its template version.

Versioned template URLs look like https://spotto-aws-onboarding.s3.amazonaws.com/cloudformation/v<version>/spotto-aws-connect.template.json. They never change once published.

Member accounts update through the StackSet. Updating SpottoMemberAccounts-<deployment> rolls the new role out to every member account, and new accounts get it automatically.

Repair​

A plain stack update doesn't fix a role someone edited in IAM, because CloudFormation compares against the last template, not the live role. Repair forces CloudFormation to reapply the trust policy and every policy.

MethodHow
CloudShellCopy the command from Repair Access, which adds --repair. In PowerShell, add -Repair.
AWS consoleUpdate SpottoConnect-<deployment> with Use current template and set RepairToken to a new number, such as today's date 20261022.
TerraformSet repair_token to a new number, then terraform apply.

What --repair does:

StepRole still existsRole was deleted
1Runs drift detection and reports what changedSame
2Updates SpottoConnect-<deployment> with a new RepairToken, reapplying the trust policy and all policiesDeletes SpottoConnect-<deployment>, recreates it with CostData=None, then deploys SpottoBillingAccess-<deployment> for the existing spotto-cur2 export
3Updates SpottoMemberAccounts-<deployment>Same

Cost data is never lost: the bucket and the spotto-cur2 export are kept when the stack is deleted, so the recreated role reads the same export and history.

Check​

--check (PowerShell: -Check) changes nothing. It prints:

ItemShows
Template versionDeployed vs latest, for each stack
DriftWhether the role or its policies differ from the template
Member accountsStackSet instances: how many are OK, outdated or failed

It exits with code 0 only when everything is current and in sync, so you can run it on a schedule.

Turn On Optional Capabilities​

Optional capabilities will be template parameters you turn on with an update, using --enable <capability>. Everything else keeps its previous value. None exist yet; the portal will show the command when one does.

A failed regional registration can leave a retained bucket before the export stack exists. The command can recreate access without recreating that bucket, grant its exact billing prefix and create the missing definition. It checks the retained bucket's account owner, region and existing delivery policy before deleting the failed access stack. A failed check stops without deletion; restore the managed policy or review the retained resources explicitly. It never empties/deletes a bucket.