November 20, 2026

ERP Disaster Recovery: What Happens If Your Hosting Provider Has a Security Incident?

Key Takeaways

  • Your ERP is only as secure as its hosting environment. A security incident at the hosting provider can disrupt ERP operations, expose critical systems, or delay business processes.
  • Third-party security and operational risks are growing. Verizon’s 2025 DBIR found third-party involvement in 30% of breaches, highlighting the importance of evaluating your hosting provider’s security practices and infrastructure.
  • Dedicated tenancy limits exposure. Separate accounts, networks, and management environments help prevent incidents affecting one customer from spreading to another. Provider access should also be restricted, protected by MFA, and governed by clear contractual requirements.
  • An ERP disaster recovery plan is only reliable when recovery has been tested. Documented runbooks, measured RTO and RPO, failover drills, and post-restore validation are essential to confirm that ERP systems can actually recover.
  • Operational readiness turns a DR plan into evidence. Runbooks, rehearsed failover, validated restores, and named owners show that recovery works before an incident forces the question.
  • Recovery plans must account for provider-side incidents. Your organization should be able to execute its recovery strategy even when the hosting provider’s consoles, accounts, or support teams are unavailable.
  • Provider accountability must be documented in the contract. Incident notification timelines, Severity 1 response targets, administrative access controls, root cause analysis (RCA), and disaster recovery commitments should be clearly defined.

Key Terms

  • Hosting provider: the company that runs the servers, storage, and network your ERP sits on, either in its own data centers or in a public cloud account it manages for you.
  • Managed service provider (MSP): a company that operates part of your IT environment under contract, such as the database, the application, or the infrastructure.
  • Shared tenancy: several customers run on the same infrastructure, accounts, or management layer.
  • Dedicated tenancy: your environment runs in its own isolated accounts and network, separate from other customers.
  • Disaster recovery (DR): the documented and tested process for bringing an ERP back at a secondary site after the primary site is lost.
  • RTO and RPO: the recovery time objective is how long the business can wait for the ERP to come back. The recovery point objective is how much data, measured in time, the business can afford to lose.
  • Root cause analysis (RCA): a written account of what failed, why, and what changed so it does not repeat.

What happens to an ERP when the hosting provider has a security incident?

Your ERP is the engine behind everyday operations. It handles closing the books, running payroll, and shipping orders. But because it relies on underlying application servers, databases, storage, and networking, your hosting provider holds the keys to the kingdom. If security breaks down at that foundational level, your ERP gets pulled into the crossfire, even if no one ever accesses your data directly.

This exposure is common. The Verizon 2025 Data Breach Investigations Report found third-party involvement in 30% of breaches, up from 15% the year before. Uptime Institute reports that third-party IT and data center providers, including cloud, telecom, and colocation, have consistently accounted for a majority of publicly reported outages over 10 years.

Downtime is priced by the hour. ITIC’s 2024 survey of more than 1,000 firms found that over 90% of mid-size and large enterprises put one hour of downtime above $300,000. For an ERP, that cost lands on month-end close, payroll runs, supplier payments, and order fulfillment at the same time.

Why does your recovery depend on how the provider is set up?

CISA and its international partners published advisory AA22-131A on threats to managed service providers and their customers. Its customer guidance reads like a checklist for the questions most contracts never answer:

  • Provider accounts should stay out of your internal administrator groups and be limited to the systems the provider manages.
  • Contracts should require multi-factor authentication on all provider accounts that access your environment.
  • Provider connectivity should run through a dedicated VPN or equivalent, separated from the rest of your network.
  • Contracts should state that the provider will not reuse administrator credentials across customers.
  • Contracts should “detail how and when MSPs notify the customer of an incident affecting the customer’s environment.”

Tenancy decides how far an incident travels. When customers share accounts, a management layer, or infrastructure, an event that starts with one tenant can reach its neighbors, and you cannot test isolation you cannot see. Under dedicated tenancy, your environment runs in its own accounts and network. At ITC, dedicated tenancy is the standard for every customer deployment.

What does operational readiness mean for an ERP, and why does it matter?

Most teams assume having a DR plan in place means they are covered, but the reality is much more complicated. Veeam’s Data Trust and Resilience Report 2026, which surveyed over 900 senior IT, security, and risk leaders, revealed that while 90% feel confident they can recover within their RTOs, and 69% believe those RTOs align with business continuity goals, the actual outcomes tell a different story. Among ransomware victims whose operations or data were impacted, only 28% managed to recover everything, with the average recovery standing at just 72% of affected data. As Veeam’s CEO clearly put it: “confidence in recovery and proof of recovery are fundamentally different capabilities.”

ERP recovery shows why. Oracle’s reference architecture for JD Edwards EnterpriseOne disaster recovery lists configuration files that must be pointed at the DR site during failover, including sqlnet.ora, tnsnames.ora, jde.ini, and jdbj.ini, and notes that the load balancer must be created manually in the standby region in advance. The same page describes start and stop drills for non-disruptive DR testing. Details like these sit in runbooks, and only a drill finds the ones that are missing.

A provider-side incident adds a second dependency. The plan has to work when the provider’s own consoles, accounts, and staff are restricted or busy, and a plan written on the assumption that they are available on day one has not been tested for that case.

ITC builds, documents, and tests DR plans for customer environments at least once a year, with post-failover checks and post-restore validation as part of the exercise.

What does operational readiness mean for an ERP, and why does it matter?

Operational readiness means your team and your provider can restore the ERP, and can show evidence that they have done it, before an incident forces the question. It rests on four things: a runbook that covers the whole environment, a failover that has been rehearsed, a restore that has been validated, and a named owner for every step.

Veeam’s 2026 research points the same way. Of the four practices the report links to stronger recovery outcomes, one is “proven recovery through realistic testing and validation,” and another is “executive alignment on ownership, reporting, and ‘what recovered means.’” The report’s own summary is that resilience improves when “readiness becomes measurable” (Veeam Data Trust and Resilience Report 2026).

In an ERP environment, readiness looks like this:

  • Runbooks: failover steps that cover the application, the database, integrations, interfaces, and configuration files, updated after every change.
  • Rehearsal: a DR drill at least once a year, with RTO and RPO measured instead of estimated.
  • Validation: post-failover checks and post-restore validation, including data checks and a sample of real transactions, before anyone declares the ERP recovered.
  • Ownership: named contacts on both sides, a Severity 1 response target, a written RCA after every incident, and a monthly review.
  • Isolation: dedicated tenancy, so an incident at another customer stays that customer’s incident.

The advantages show up in four places:

  • Predictable recovery: the team runs the steps for the first time in a drill, so the missing load balancer or the stale configuration file turns up while the business is still running.
  • A shared definition of recovered: finance, operations, and IT agree in advance on what recovered means, such as a completed test close and reconciled balances, so the ERP is not handed back on the strength of a login screen.
  • Fewer ownership disputes: the contract and the runbook name who acts, so the first hour of an incident goes to recovery.
  • Evidence for leadership: dated test reports and measured RTO and RPO give executives and auditors something to review besides reassurance.

At ITC, operational readiness is built into the service. DR plans are built, documented, and tested at least once a year, with post-failover checks and post-restore validation, and RCA and monthly reviews are standard practice.

What should a stable provider put in writing?

A stable provider answers each of these with a document, a date, or a number.

Area Ask for What a good answer looks like
Administrative access Who at the provider can reach your environment, and through what path Named roles, MFA on every provider account, a dedicated access path, and accounts disabled when unused
Tenancy Whether your accounts and network are separate from other customers Dedicated tenancy stated in the contract
Disaster recovery The date and result of the last end-to-end test A runbook, a dated test report, and RTO and RPO figures that were measured
Incident notice How fast, through which channel, and with what content Notification terms written into the contract
Support response The Severity 1 response target A stated target
After an incident The RCA and the review cadence A written RCA for every incident and a recurring review

What changes when your provider is acquired or reorganized?

Providers change hands, and the contract usually follows. The people, tools, and escalation paths behind it can change over time. Four checks show whether that has happened to your account:

  1. Who is your named escalation contact today, and was it the same person a year ago?
  2. Has the support model, location, or tooling changed since you signed?
  3. Are your DR runbooks dated after the last change?
  4. When did the provider last send a written review of your environment?

The same checks apply if you moved to a new provider recently. A new contract with a provider whose recovery has never been tested carries the same exposure as the old one, and the questions cost nothing to ask.

Six questions to answer about your own provider

Answer each one with a document or a date. Any answer that starts with “I think so” is a question for your provider this month.

  1. Can you name every provider account that can reach your ERP, and is multi-factor authentication enforced on each of them?
  2. Do you know whether any other customer shares your accounts, network, or management layer?
  3. When was your last DR test that included failing over, validating data, and running real transactions, and what recovery time did it measure?
  4. Would your DR plan work if the provider’s consoles and staff were unavailable for the first day?
  5. Does your contract say how quickly the provider must tell you about an incident, and who will make that call?
  6. After the last major incident, did you receive a written root cause and a list of changes?

Talk to an ITC specialist

ITC manages Oracle environments under dedicated tenancy, with a 15-minute Severity 1 response target, annual DR testing, written root cause analysis, and monthly reviews. If you want a second opinion on how your current setup answers the six questions above, book a 15-minute conversation with an ITC specialist.

Frequently Asked Questions (FAQs)

  1. What should we do first when our hosting provider reports a security incident?
    Ask for a written statement of which systems and accounts are affected, and confirm the current state of provider access to your environment. Then open your DR runbook and check your own logs. CISA advises customers to enable monitoring and logging of their own systems, so you have a record that does not depend on the provider (CISA AA22-131A).
  2. Who is responsible for recovery, the provider or us?
    The contract decides. CISA recommends that contracts specify whether the provider or the customer owns specific responsibilities such as hardening, detection, and incident response (CISA AA22-131A). If the contract says nothing, you will find out who owns recovery in the middle of the incident.
  3. Is dedicated tenancy more secure than shared tenancy?
    Dedicated tenancy limits how far an incident at another customer can travel, because your accounts and network are separate. It does not replace access controls, monitoring, or tested recovery, which apply under either model.
  4. How often should we test ERP disaster recovery?
    At least once a year, and after major changes to the environment. ITC tests customer DR plans at least annually. Oracle’s JD Edwards DR guidance includes start and stop drills so the test does not interrupt production (Oracle documentation).
  5. What is the difference between RTO and RPO?
    RTO is how long the business can wait for the ERP to come back. RPO is how much data, measured in time, the business can afford to lose. A plan needs both numbers, and both need to be measured in a test.
Debra Castillo
Debra Castillo Debra has over 11 years of experience working across Cloud Infrastructure, Managed Services, Workload Migrations, Tariff Impact Management, and Nearshoring. She specializes in writing compelling narratives to address complex challenges.

Related Blogs