SERVICES · LAS VEGAS NV

Charles Trinh — Multi-Tenant Single-Host Design | Las Vegas, NV · Las Vegas, Nevada · Henderson · Clark County, Nevada · North Las Vegas, Nevada

Charles Trinh — Multi-Tenant Single-Host Design, Las Vegas NV
Charles Trinh — Multi-Tenant Single-Host Design

Running several tenants on one host without leaking between them: isolation choices, data partitioning, noisy-neighbour control, and the failure that crosses a boundary.

Isolation is a spectrum

Separate schemas, separate databases, separate processes and separate hosts differ in blast radius and in cost. Choose by what a boundary breach would cost, not by what is easiest to deploy.

Document the choice explicitly, because the next engineer will assume the stronger option than what is actually running.

Tenant identity comes from the session

Resolve tenant identity from the authenticated session and enforce it at the data layer. A tenant parameter in a request is a suggestion, not a control.

Test cross-tenant access with a request that deliberately tries to read another tenant, and keep that test in the pipeline.

Partition the things you restore

Partition storage and backups per tenant so one restore does not touch another. A shared backup means a per-tenant recovery is impossible exactly when it is needed.

Keep per-tenant record counts and verify them after any restore, since a silent shortfall is worse than a failure.

Bound the noisy neighbour

Set CPU, memory and connection ceilings per tenant, because the ordinary failure in a shared host is one tenant consuming the estate.

Measure per tenant and alert on the tenant, not on the host, so the first notification names the cause.

Choosing the isolation model

The options run from a tenant column in a shared table to a host per tenant, and each step up costs money and buys blast radius. The decision should follow from what a cross-tenant exposure would cost, which is a question for the business rather than the engineering team. Where the answer is severe, the design should be the stronger option even when the traffic does not justify it, because the justification for weak isolation is always cost and the cost of a breach is not comparable.

Document the choice and the reasoning, and make the document part of the architecture review. The next engineer will assume a stronger model than the code implements, and the assumption is invisible until an incident tests it.

Tenant identity from the session

Tenant identity must come from the authenticated session and be enforced at the data layer. A tenant identifier in a request body or a query parameter is a suggestion, and a client that changes it is not attacking so much as testing. The enforcement belongs in the query construction rather than in the handler, because a handler that checks and a query that does not is a bypass with a comment explaining the check.

Test cross-tenant access with a deliberate request in the pipeline, and keep the test as a regression. The test that tries to read another tenant is the cheapest assurance in a multi-tenant system and it is the one most often missing.

Partition what you restore

Backups, exports and restores should be partitionable by tenant, because a shared backup makes a per-tenant recovery impossible exactly when a tenant needs one. Keep per-tenant record counts and verify them after restore, since a silent shortfall is worse than a visible failure and it surfaces weeks later as a data quality complaint.

Where a shared backup is unavoidable, document the per-tenant extraction procedure and test it once. The test is what turns a theoretical capability into a fact.

Noisy neighbour controls

Set CPU, memory and connection ceilings per tenant, because the ordinary failure in a shared host is one tenant consuming the estate rather than an attack. Measure per tenant and alert on the tenant, so the first notification names the cause rather than the host. Ceilings that are never hit are cheap; ceilings that are discovered during an incident are a redesign.

Where a tenant legitimately needs more, make it a documented allocation rather than a temporary exception. Temporary exceptions become permanent and the model loses its meaning.

Schema changes in a shared host

A schema change affects every tenant at once, which makes the migration a release with a blast radius. Test against a dataset that includes the awkward tenants, keep the migration reversible, and schedule it with the same care as any other change with a wide effect. The convenience of one schema is real and it is paid for at every migration.

What the design costs at scale

A shared host gets cheaper with every tenant until the first tenant that needs isolation for contractual reasons, and from then on the estate has two designs and one of them is under-exercised. Plan for the exit at the start: the ability to move a tenant to a dedicated host with a documented procedure and a tested run, even if it is never used. That capability is what lets isolation be sold as an option rather than promised as a principle.

Measure the cost per tenant including the shared overhead, because a pricing model built on the marginal cost understates the fixed components that do not shrink. Tenants that leave take their marginal cost with them and leave the fixed cost behind, and a model that does not account for that discovers it at the worst time.

Revisit the isolation choice when the tenant base doubles, or when a tenant asks. Those are the two moments when the answer could legitimately change.

Incident isolation

When something goes wrong, the first question is which tenants are affected, and the answer has to be available in minutes. Keep per-tenant logging and per-tenant metrics from the beginning, because retrofitting them during an incident is impossible. The capability to say which tenant a problem started with is the difference between a contained incident and a general outage.

Rehearse one tenant-scoped incident per year: a tenant whose data is wrong, not the host. The rehearsal finds the places where the system assumes one tenant, and those places are the design's real weak points.

When a tenant leaves

Offboarding a tenant is a data deletion exercise with an obligation attached, and it should be a documented procedure rather than a support task. The procedure covers the export, the confirmation to the tenant, the deletion across primary storage, backups and the search index, and the record of when each happened. A tenant that leaves without a deletion record is a retention liability that nobody tracked.

Where backups must be retained, describe the retention in the confirmation and state the date after which the data is unrecoverable. That sentence is what a former tenant needs and what your own counsel will ask for.

Keep the deletion evidence with the tenant record, since the question after a departure is always what was removed and when.

Keep per-tenant rate limits in the same configuration as the isolation model, so a tenant's ceiling is a documented allocation rather than a value somebody edited during an incident.

Record the isolation decision with a date and a reviewer, so the next review starts from a position rather than from a blank page.

Review the tenant list quarterly against the billing system, since a tenant that exists in one and not the other is either revenue leakage or an unowned environment.

Proof

  • ONC (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — public suite URL
  • 67 public hostnames, 51 serving, one AWS account
  • 104 clearinghouse enrolments across 53 jurisdictions
  • USCDI v7 proposal — Patient Digital Identity Credential — listed at isp.healthit.gov
  • SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564
  • AWS Business Associate Addendum active since 18 August 2026

Where this work is delivered

Engagements run from Las Vegas and reach across the valley: clients in Las Vegas, NV; Las Vegas, Nevada; Henderson; Clark County, Nevada; and North Las Vegas, Nevada — plus remote delivery nationwide.

How an engagement runs

1. Trace one real case

Follow a single patient or claim end to end with a timestamp at every hop, before any document is written.

2. Written findings

A findings list naming the defect, the system, the owner and the date.

3. Monthly review

Review the list monthly and check whether the dates held.

Frequently asked

What is this service?

This page describes charles trinh — multi-tenant single-host design — the scope, the failure modes it prevents and how the work is measured. It sits in the Cloud and Operations category at trinh.io.

Where is the work delivered?

Engagements run from Las Vegas, Nevada across the valley — Las Vegas, Henderson, North Las Vegas, Spring Valley, Paradise, Winchester, Sunrise Manor, Enterprise, Summerlin, Clark County — and remotely anywhere in the United States. On-site work in the Las Vegas valley, remote delivery elsewhere.

What does it cost?

Standard rate $195 per hour, committed $165 per hour at twenty hours or more per week for three months or more; $1,600 standard day, $1,350 committed day. Scoped in writing before work starts, fixed fee where the deliverable is defined.

Who is Charles Trinh?

A technology architect in Las Vegas, Nevada, working in healthcare identity, consent and interoperability. Not a physician, not a medical practice, and not affiliated with any other Charles Trinh.

How does an engagement start?

With one trace of a real case and a written findings list naming the defect, the system, the owner and the date. The first call is twenty minutes and free: 702-559-3489.

Engagement

Scoped in writing before work starts; fixed fee where the deliverable is defined. First call is 20 minutes and free. 702-559-3489 · charles@trinh.io

Talk to Charles

Record

Profile

Healthcare technology architect working where federal regulation becomes production infrastructure. Built and operates a FHIR R4 clinical platform — SMART App Launch authorization server, patient and provider portals, SQLite — on a hardened single-host AWS estate of 67 public hostnames. The platform returned a self-test PASS on the ONC (g)(10) Standardized API test suite in April 2026, 317 of 317 subtests, publicly verifiable. Designed a consent architecture in which a permissioned ledger records patient grant, refusal and revocation as immutable events while all clinical data stays off-chain and encrypted per patient — no PHI on the ledger by design. Filed a proposed USCDI v7 data element with ONC on verifiable patient identity. Translates the Federal Register, HHS/CMS guidance and HITRUST requirements into architecture, controls and audit evidence.

Selected engineering scale

  • Self-managed estate of 67 public hostnames, 51 serving, on one AWS account.
  • FHIR R4 API with a SMART App Launch authorization server; patient and provider portals; an AI service layer; Hyperledger Fabric peer, orderer and CA; IPFS; a 337-table SQLite clinical store.
  • Two permissioned distributed ledgers in production, one operating a consent registry that records grant, refusal and revocation with no PHI on chain.
  • X12 005010 pipelines across eight transaction sets.
  • Designed, built, deployed and operated end to end by one person.

Credentials and standards work

  • Inferno (g)(10) self-test PASS, 317 of 317 subtests, April 2026 — the ONC Standardized API test suite. Standards validated: US Core 6.1.0, USCDI v3, SMART App Launch 2.0.0, Bulk Data 2.0.0. Public suite URL.
  • USCDI v7 public comment submission, April 2026 — proposed Patient Digital Identity Credential under the Provenance data class, listed publicly on the ONC Interoperability Standards Platform.
  • SAM.gov UEI MVKQW9DKKGB8 · CAGE 20KM5 · NPI 1194645564. AWS Business Associate Addendum active 18 August 2026.

Technical environment

AWS us-east-1 (compute, storage, messaging, SMS, email, voice) · Linux and systemd hardening · nginx · Fabric and EVM ledgers · FHIR R4 · X12 005010 · HL7 · SQLite · self-hosted mail, drive, search and monitoring · Git-based change control and CI release gates.

TRINH — Charles Trinh, technology architect
TRINH

Charles Trinh

Fractional CTO/CIO and Technology Architect — Healthcare Data, Identity & Regulated Infrastructure · Las Vegas, NV

Las Vegas, Nevada · US citizen · Fractional and project engagements

(702) 559-3489 Email Charles Trinh