# Hosting Security: 12 Essential Checks for Safer Websites

Source: https://hostreta.com/2026/10/hosting-security-checks/

In this guide: Is your web hosting secure? 12 security features to check — server-level protections, account isolation, backups and recovery, and questions to ask your host.

Website security doesn’t start with a WordPress plugin. It starts with your hosting. If the server is outdated, poorly isolated, or unmonitored, even a perfectly maintained website is at risk.

Use these twelve features as questions to verify for your actual plan, not as a claim that one product name guarantees complete protection.

Table of Contents

[Toggle](#)

- [Why hosting security matters](#Why_hosting_security_matters)
- [Server-level hosting security protections](#Server-level_hosting_security_protections)
- [Account isolation in hosting security](#Account_isolation_in_hosting_security)
- [Backups and recovery](#Backups_and_recovery)
- [Hosting security questions to ask your provider](#Hosting_security_questions_to_ask_your_provider)
- [Your hosting security responsibilities](#Your_hosting_security_responsibilities)
- [Key Takeaways: Hosting Secure](#Key_Takeaways_Hosting_Secure)
- [Hosting security evidence: start with responsibilities](#Hosting_security_evidence_start_with_responsibilities)
- [1. Verify the request-filtering boundary](#1_Verify_the_request-filtering_boundary)
- [2. Establish DDoS coverage for the actual service](#2_Establish_DDoS_coverage_for_the_actual_service)
- [3. Check scanner coverage and response](#3_Check_scanner_coverage_and_response)
- [4. Verify runtime support and patch responsibility](#4_Verify_runtime_support_and_patch_responsibility)
- [5. Review authentication for infrastructure accounts](#5_Review_authentication_for_infrastructure_accounts)
- [6. Treat TLS as transport protection](#6_Treat_TLS_as_transport_protection)
- [7. Understand account and site isolation](#7_Understand_account_and_site_isolation)
- [8. Distinguish resource controls from guaranteed capacity](#8_Distinguish_resource_controls_from_guaranteed_capacity)
- [9. Keep control-plane permissions reviewable](#9_Keep_control-plane_permissions_reviewable)
- [10. Define usable backup requirements](#10_Define_usable_backup_requirements)
- [11. Confirm restoration and data-loss consequences](#11_Confirm_restoration_and_data-loss_consequences)
- [12. Make incident escalation operational](#12_Make_incident_escalation_operational)
- [Services to compare against your requirements](#Services_to_compare_against_your_requirements)
- [Illustrative hosting security evidence record](#Illustrative_hosting_security_evidence_record)

## Why hosting security matters

Your host controls the layers beneath your website:

- The server operating system and its security patches.

- The web server, PHP, and database software.

- The network and firewall.

- Isolation between customers on shared servers.

- Backups and disaster recovery.

- Monitoring for attacks and malware.

Hosting controls can reduce particular risks and support recovery, but their effectiveness depends on coverage, configuration and the division of responsibilities.

## Server-level hosting security protections

1. Web Application Firewall (WAF)

Filters malicious requests (SQL injection, cross-site scripting, known exploit attempts) at the server level. See [Web Application Firewalls Explained](https://hostreta.com/2026/10/waf-guide/).

2. DDoS protection

Network-level mitigation to absorb traffic floods. Many hosts partner with CDNs for this. See [DDoS Attacks Explained](https://hostreta.com/2026/10/ddos-response-guide/).

3. Malware scanning

Automatic scanning of your files for known malware, ideally with alerts and cleanup options.

4. Up-to-date software stack

Current, supported versions of PHP, MySQL/MariaDB, and the operating system, with security patches applied promptly. The host should let you choose supported PHP versions and warn about outdated ones.

5. Brute-force protection

Rate limiting or blocking repeated failed logins to control panels, SSH, FTP, and common CMS login pages.

6. Free SSL and secure protocols

Free SSL certificates with auto-renewal, TLS 1.2/1.3, and support for SFTP/SSH instead of plain FTP.

## Account isolation in hosting security

7. Isolated accounts on shared servers

Technologies like CloudLinux (with CageFS), containers, or separate system users keep your account’s files and processes separated from other customers. Without isolation, one compromised site could affect neighbors. See [What Is Server Isolation?](https://hostreta.com/2026/10/server-isolation/).

8. Resource limits

Per-account resource controls can reduce noisy-neighbor effects, but do not guarantee availability or eliminate every shared dependency.

9. Secure control panel access

Two-factor authentication for your hosting account and control panel, activity logs, and the ability to create limited-access users for developers.

## Backups and recovery

10. Automatic backups

Confirm frequency, protected components, independent storage and retention against the amount of data the business can afford to lose.

11. Easy restore

One-click restore of files, databases, or the whole account, preferably free of charge. Ask how long restores take.

12. Monitoring and incident response

24/7 infrastructure monitoring, a public status page, and a clear process for handling hacked accounts — including whether malware cleanup is included or paid.

Remember the [3-2-1 backup rule](https://hostreta.com/2026/09/website-backups-the-3-2-1-rule-explained/): keep your own off-site backup too.

## Hosting security questions to ask your provider

Use these before buying — or ask your current host today:

- Do you run a web application firewall? Is it included?

- How are accounts isolated on shared servers?

- How often are backups taken, where are they stored, and how long are they kept?

- Is restoring a backup free?

- Do you scan for malware, and is cleanup included?

- Which PHP versions do you support, and how do you handle end-of-life versions?

- Do you support 2FA for the hosting account?

- Do you offer SSH/SFTP access (and disable plain FTP)?

- What DDoS protection is in place?

- What happens if my site is hacked — what support do you provide?

Clear, confident answers are a good sign. Vague answers are a red flag.

## Your hosting security responsibilities

Even the best host can’t protect you from:

- Weak or reused passwords.

- Outdated plugins and themes.

- Nulled (pirated) software.

- Giving admin access to people who don’t need it.

Use the [WordPress Security Checklist](https://hostreta.com/2026/10/wordpress-security-checklist/).

Further reading: [WordPress Advanced Administration Handbook](https://developer.wordpress.org/advanced-administration/)

## Key Takeaways: Hosting Secure

- Hosting security is the foundation for website security.

- Look for WAF, DDoS protection, malware scanning, and updated software.

- Account isolation protects you from compromised neighbors.

- Verify backup freshness, retention, restore access and any assistance charges against your business requirements.

- Ask your host direct questions — and do your part too.

## Hosting security evidence: start with responsibilities

Ask the provider to explain the responsibilities for your actual service, not a general product category. Shared hosting, managed WordPress and an unmanaged VPS can place very different work on the customer. The label managed is not a complete contract. Record who updates the operating system, runtime, application and plugins, who responds to alerts and who authorizes a recovery operation.

The [Australian government’s guidance for small businesses](https://www.cyber.gov.au/publication/cloud-shared-responsibility-model-guidance-for-individuals-and-small-and-medium-businesses) explains why responsibility remains shared when using cloud services. Apply that principle to the exact hosting plan and written service scope. Hosting security questions should produce an accountable owner and a verifiable control, rather than a vague assurance that the provider handles everything.

### 1. Verify the request-filtering boundary

A firewall may operate at the edge, on the server or inside the application. Identify where it evaluates requests and which hostnames are covered. A security plugin can be useful without protecting the upstream network from a flood. A remote service may not inspect direct requests to an unprotected origin. Keep the actual request path documented before assuming that a feature is active.

For hosting security, ask who maintains rules and investigates false positives. A provider should explain how legitimate forms, APIs and callbacks are validated. The [WAF workflow](https://hostreta.com/2026/10/waf-guide/) describes narrow exceptions and a controlled rollout. Do not measure protection by the number of blocked requests alone or treat every event as a confirmed prevented breach.

### 2. Establish DDoS coverage for the actual service

Ask which protocols and services the mitigation covers and whether the relevant hostname is onboarded correctly. Network and HTTP protection are related but different. Confirm support escalation and any actions the business owner must authorize during an incident. A general statement that a provider has a large network does not explain what happens to your particular account.

Keep hosting security and capacity planning distinct. A bigger server may help legitimate workload demand without solving a saturated network path. Review the [DDoS response guide](https://hostreta.com/2026/10/ddos-response-guide/) and preserve a status channel outside the affected site. Do not promise uninterrupted availability or describe an unfamiliar traffic source as malicious without relevant evidence.

### 3. Check scanner coverage and response

Ask whether scanning examines public responses, local files, databases or some combination, and which exclusions apply. Find out whether a finding produces an alert, automatic quarantine or manual support work. A completed scan does not prove that every component is clean. A vulnerability warning also does not establish that exploitation has occurred.

In a hosting security review, give each finding an owner and a defined next step. Confirm whether cleanup is included, optional or outside the plan. Read the [malware scan guide](https://hostreta.com/2026/10/malware-scan/) before approving bulk repair. Protect evidence and legitimate custom code instead of assuming that everything unfamiliar should be deleted automatically.

### 4. Verify runtime support and patch responsibility

Check the runtime actually serving the website and the provider’s patch process. A command-line version may not describe every web worker. Consult [PHP’s current support table](https://www.php.net/supported-versions.php) and ask the provider about any distribution or vendor backports rather than judging solely by a shortened version string. Keep compatibility testing separate from the need to address security updates.

Hosting security includes a plan for end-of-life software. Ask how customers are warned, how supported upgrades are tested and who handles application incompatibility. Do not stay on an unsupported branch indefinitely merely because the site still loads. Conversely, do not switch a production runtime without appropriate validation and recovery access. Record the actual supported path and responsible owner.

### 5. Review authentication for infrastructure accounts

Examine the hosting portal, control panel, SSH access and recovery mailbox separately. They may have different identities, permissions and second-factor capabilities. Individual developer access is easier to review than one shared administrator password where the platform supports delegation. Keep recovery methods current and document how the business retains control when a contractor leaves.

The [credential ownership guide](https://hostreta.com/2026/10/password-manager-guide/) supports this part of hosting security. Removing one panel user may not revoke an API token, deployment key or separate account. Review those routes as part of offboarding. A successful login with two factors is useful evidence of that workflow, not proof that all other account recovery paths are equally protected.

### 6. Treat TLS as transport protection

A certificate helps authenticate the connection endpoint and protect transport when configured correctly. It does not prove that the business is trustworthy or that the site’s code is uncompromised. Confirm renewal ownership, supported protocol configuration and monitoring for expiry. If a reverse proxy is present, review the connection from the proxy to the origin as well as the browser-facing endpoint.

For hosting security, ask whether secure file transfer and administrator access are supported through the documented interfaces. Avoid transmitting credentials through insecure legacy workflows merely for convenience. Do not bypass a certificate warning as an ordinary troubleshooting shortcut. Have the responsible administrator investigate the actual certificate or routing problem and verify the correction.

### 7. Understand account and site isolation

Ask what boundary separates customers and whether several sites inside one customer account share credentials or file access. [CloudLinux’s component documentation](https://docs.cloudlinux.com/cloudlinuxos/cloudlinux_os_components/) describes CageFS as a virtualized filesystem isolation mechanism. That is not the same as a guarantee that every site in every account is separately isolated or that all operating-system risks disappear.

Document the configured boundary for your service and consider the business impact of a shared account compromise. Hosting security should reflect the actual arrangement, especially for agencies managing unrelated clients. Use the [isolation guide](https://hostreta.com/2026/10/server-isolation/) to distinguish filesystem, process and resource controls. A technology name without configuration evidence is an incomplete answer.

### 8. Distinguish resource controls from guaranteed capacity

CPU, memory, process and I/O limits describe different constraints. Ask how the host measures them and what happens when a limit is reached. Resource isolation can reduce interference from neighboring workloads without reserving unlimited capacity. Marketing terms such as unlimited storage or bandwidth need their actual acceptable-use and technical conditions reviewed.

Keep performance and hosting security observations specific. A resource spike may reflect a normal task, application defect or unwanted activity. Investigate before labeling it an attack. Record which metric reached a limit and which business function failed. This gives the provider a useful diagnostic report instead of an unsupported demand to remove every restriction.

### 9. Keep control-plane permissions reviewable

Record which users can manage domains, backups, databases, billing and new administrators. These privileges do not always align with the roles inside WordPress. Use supported limited-access accounts for routine work and keep ownership with the business. Confirm the process for reviewing logs and responding to an unexpected administrator change.

Hosting security reviews should include connected applications and API keys, not just visible human users. Give each integration a business owner and remove unused authorization through the supported process. The [security review record](https://hostreta.com/2026/10/website-security-review/) helps maintain evidence without exposing secrets in an ordinary shared document.

### 10. Define usable backup requirements

Ask what data is protected, how often recovery points complete, where they are stored and how long they remain available. Consider whether website compromise or account suspension could also prevent access to the copies. Match the frequency to the site’s writes and acceptable data gap. A daily copy may be inadequate for a busy store and more than some static content needs.

For hosting security, verify the latest usable recovery point rather than relying only on an enabled setting. Use the [automatic backup guide](https://hostreta.com/2026/10/automatic-backups/) to plan independent storage and controlled restoration tests. A backup is not a complete recovery capability until the authorized team understands how to use it safely.

### 11. Confirm restoration and data-loss consequences

Ask whether restore is self-service or requires support, what fees apply and how authorization works if the main panel is unavailable. A full account archive may need a different workflow from a partial file restore. Test representative recovery in an isolated environment where practical, preventing customer emails and production payments from the copy.

Rolling back a database can overwrite newer orders, registrations or messages. Keep that consequence visible when planning hosting security and incident recovery. Do not restore production casually because a button is convenient. Record the chosen point, expected gap, owner and validation results. A demonstrated exercise is an observation under those conditions, not a guaranteed production recovery time.

### 12. Make incident escalation operational

Keep emergency contacts, account identifiers and permitted actions available outside the website. Confirm how incidents are reported and who can authorize restrictions or recovery. Ask what monitoring actually checks and how alerts reach the responsible person. Around-the-clock monitoring is not the same as a promise that every application problem receives immediate manual remediation.

Coordinate customer updates through the [communication plan](https://hostreta.com/2026/10/crisis-communication/). In a hosting security incident, state confirmed facts and unresolved questions. Do not claim that data, payments or access were unaffected before the investigation supports it. Review the division of work afterward and keep unfinished actions assigned rather than closing the record as soon as the homepage returns.

### Services to compare against your requirements

Promotional recommendation: review [KAWEDMA’s hosting services](https://kawedma.com/) and ask for the controls and responsibilities of the specific plan. Its service offering is a candidate to evaluate, not a certification of performance or complete protection. Confirm backup, support, isolation and incident-response terms in current documentation.

Affiliate disclosure: Hostreta may earn a commission if you purchase through these links. Compare [hosting.com’s managed WordPress service](https://my.hosting.com/aff/08a016e41d3100e5460ff6ebee1b5e0a91e643d6) for a WordPress-focused scope, or [managed VPS](https://my.hosting.com/aff/a4acdd3ba5260a0812f3b4249da227e940d132e7) when the workload and support requirements justify it. Confirm the actual management boundary; neither name alone answers all twelve checks.

Hosting security: assign responsibility. Original explanatory diagram.

Hosting security: separate the controls. Original explanatory diagram.

Hosting security: prove operational readiness. Original explanatory diagram.

### Illustrative hosting security evidence record

For each of the twelve checks, keep the plan name, source document, date reviewed, responsible owner and the evidence still needed. Separate what the provider states from what your team has validated. For example, a plan may advertise backups, while the team has not yet checked whether it can restore the relevant database. Mark that restoration capability unverified instead of giving the entire plan a passing grade.

Consider a small store moving from an unmanaged server to a managed service. The owner should ask which operating-system and runtime tasks transfer to the provider and which application tasks remain with the business. The team should also confirm that its payment integrations, recovery routes and access roles remain functional. This is a planning example, not a measured migration or a claim about any vendor’s response time.

After a change, review the same hosting security record rather than starting from an unrelated checklist. Confirm whether the evidence still applies to the current plan, account and environment. If a provider cannot disclose a technical implementation detail, ask for the relevant customer-facing responsibility and support process rather than inventing an answer. Record the limitation and decide whether it is acceptable for the workload.

The result should be a small set of clear decisions: which controls are confirmed, which require configuration and which remain outside the purchased service. Assign an owner to the gaps and revisit them after meaningful changes. Hosting security becomes more manageable when responsibilities are explicit than when every feature is represented by an unexplained green checkmark.
