Use this web hosting security checklist to harden a new or existing site. It covers hosting-account protection, HTTPS, updates, backups, malware monitoring, permissions, firewall rules, SSH, and incident response so you can turn security into a repeatable operating routine rather than a one-time setup task.
Overview
Secure web hosting is a shared responsibility. Your provider may protect the physical infrastructure, virtualization layer, network, or managed server components, but you still control important parts of the website: account access, applications, plugins, databases, files, DNS, and recovery procedures. The exact controls vary between shared hosting, managed WordPress hosting, and a self-managed VPS, so treat this checklist as a framework and confirm implementation details in your provider’s documentation.
Start by identifying what you are protecting and who needs access. Record each domain, application, database, hosting account, administrator, deployment method, and third-party integration. Then apply controls in this order:
- Protect the hosting account and domain-management accounts.
- Reduce the number and privilege level of users.
- Encrypt traffic and remove insecure administrative paths.
- Keep the operating system, control panel, runtime, CMS, themes, and extensions maintained.
- Create independent backups and verify that they can be restored.
- Monitor for suspicious activity and document an incident response plan.
If you are deciding whether a shared plan or VPS gives you the control you need, compare the administration responsibilities before moving. Our guide to shared hosting versus VPS hosting can help frame that decision. A VPS gives you more configuration authority, but it also makes patching, firewall management, SSH security, and recovery more dependent on your own process unless the server is managed.
Checklist by scenario
For every hosting account
- Enable MFA: Turn on multi-factor authentication for the hosting control panel, billing account, domain registrar, primary email account, and any cloud or deployment service connected to the site. Store recovery codes in a protected password manager or other controlled location.
- Use unique credentials: Do not reuse the hosting password elsewhere. Prefer a long, randomly generated password and review whether the provider supports passkeys or other stronger authentication methods.
- Audit users and access: Remove former collaborators, unused FTP accounts, old API tokens, abandoned control-panel users, and integrations that no longer serve a purpose. Give each person an individual account where possible so access can be reviewed and revoked without sharing credentials.
- Protect the email account: Hosting resets commonly depend on email. Secure that mailbox with MFA, unique credentials, recovery options, and a review of forwarding rules.
- Review domain and DNS access: Lock down the registrar account, check nameserver changes, and document which DNS records are required. Unrecognized DNS changes can redirect traffic or interfere with email even when the web server itself is intact.
- Check account notifications: Enable alerts for password changes, new logins, billing changes, account recovery, and other security events when available.
For a website or CMS
- Use HTTPS everywhere: Install a valid certificate, redirect HTTP to HTTPS, and update application URLs and canonical settings consistently. Test login forms, checkout pages, APIs, images, scripts, and webhooks for mixed-content or certificate errors.
- Keep software current: Maintain the CMS, themes, plugins, libraries, PHP or other runtime, control panel, and server packages. Remove software that is unused rather than leaving it disabled but exposed.
- Limit administrator accounts: Use separate accounts for administration and ordinary work. Assign the lowest role that completes the task, and avoid using a superuser for routine publishing or deployment.
- Protect configuration files: Keep database passwords, API keys, and environment files outside publicly served directories when the application supports it. Never commit secrets to a public repository or paste them into support tickets.
- Review file permissions: Make files readable and writable only where required. Avoid making an entire site directory writable by the web server unless the application genuinely needs that access. Confirm ownership and permissions after migrations or automated deployments.
- Use malware scanning carefully: Schedule scans if the host provides them, but treat a clean scan as one signal rather than proof that the site is safe. Investigate unexpected files, administrator accounts, redirects, modified timestamps, and outbound mail.
For a VPS or dedicated server
- Patch the operating system: Establish a regular update process and know which service restarts may affect availability. Test significant changes before applying them to a production system.
- Harden SSH: Prefer key-based authentication, restrict SSH to trusted addresses or a controlled access path where practical, disable direct root login when your administration model permits it, and remove unused keys. Keep a tested recovery route before changing SSH settings.
- Configure a host firewall: Allow only required services, such as web traffic and a restricted administrative path. Do not open database, cache, control-panel, or internal service ports to the public internet unless there is a documented need and compensating protection.
- Separate services: Use distinct system users and permissions for applications where practical. Avoid running every process as root.
- Review logs: Monitor authentication attempts, privilege changes, web-server errors, firewall events, scheduled tasks, and application logs. Send important logs to a location that an intruder cannot easily alter.
- Document changes: Record firewall rules, installed services, deployment steps, backup locations, and recovery commands. Documentation reduces mistakes during an outage or security event.
Managed VPS hosting can reduce routine administration, but it does not remove the need to understand who handles patching, backups, firewall changes, monitoring, and incident response. Confirm the division of responsibility in writing. For context, see our comparison of managed and unmanaged VPS hosting.
Backup verification worksheet
Do not mark backups complete merely because a job reports success. Record the following for each important site:
- Backup scope: files, databases, email, DNS records, certificates, configuration, and application secrets.
- Backup location: provider storage, separate storage, or another controlled destination.
- Schedule and retention: how often backups run and how many recovery points are retained.
- Protection: account access, encryption where appropriate, and whether backups are isolated from the production account.
- Last restore test: date, restored components, test result, and time required.
- Recovery owner: the person who can start restoration and the person who approves production changes.
Perform a restore to a temporary environment, not directly over production, whenever possible. Confirm that the site loads, users can authenticate, databases are consistent, uploaded files are present, scheduled tasks work, and DNS or certificate steps are understood. Provider retention and backup scope can change, so verify them instead of assuming that a plan includes every component you need.
What to double-check
Before calling a site hardened, review these areas manually:
- Control-panel settings: Confirm MFA, login alerts, backup configuration, malware scanning, SSL management, PHP or runtime selection, account-user controls, and access logs. Names and locations differ between panels, so follow the host’s current documentation.
- TLS settings: Use the secure protocol versions and cipher configuration supported by your host and application. Disable obsolete options when the provider permits it, but test compatibility with older clients, APIs, payment services, and monitoring tools.
- Permissions after deployment: Check that deployment did not expose a repository, environment file, backup archive, debug log, database dump, or temporary upload directory.
- Publicly reachable services: Scan your own documented public surface and compare it with the services you intended to expose. Close anything unnecessary.
- Recovery access: Verify that at least two authorized people can reach the password manager, backup location, hosting account, and emergency documentation without sharing a single personal credential.
- Monitoring and uptime: Use external availability checks and application-level checks where appropriate. A server can respond while a login, checkout, database connection, or background job is failing. Our web hosting performance test guide explains how to evaluate response time and availability as part of a broader review.
Common mistakes
- Relying on the host’s security label: “Secure hosting” does not describe every setting on your account. Ask what is managed, what is excluded, and how incidents are communicated.
- Leaving old access active: Temporary migration accounts, contractor logins, staging credentials, and forgotten SSH keys often outlive their original purpose.
- Treating backups as an archive: A backup that cannot be restored, is stored under the same compromised account, or excludes the database may not support recovery.
- Changing firewall or SSH rules without a fallback: Keep a console or provider recovery route available and make one change at a time so an error can be reversed.
- Installing overlapping security tools blindly: Scanners, caching layers, firewalls, and plugins can conflict or create false confidence. Document what each tool detects and who reviews its alerts.
- Ignoring staging and subdomains: Test sites often contain copied production data and weaker credentials. Protect them, remove sensitive data, and prevent search indexing when appropriate.
- Failing to preserve evidence: If compromise is suspected, avoid immediately deleting files or logs. Record times, alerts, affected accounts, and relevant copies before rebuilding or cleaning the system.
If a site is compromised, isolate affected accounts or services where practical, preserve useful logs, rotate credentials from a trusted device, identify the initial access path, restore from a known-clean backup, patch the weakness, and monitor for recurrence. Contact the host when infrastructure or account access may be involved. For availability incidents, use a separate website outage response checklist so security investigation and uptime communication are handled together.
When to revisit
Run this checklist during initial onboarding, before a seasonal traffic period, after adding a collaborator, before launching a new application, and after a migration. Schedule a shorter monthly review for users, MFA, alerts, updates, backups, malware findings, and unusual logs. Perform a deeper quarterly review of firewall rules, exposed services, TLS compatibility, permissions, DNS, recovery documentation, and restore testing.
Update the checklist whenever your provider changes its control panel, supported runtime versions, TLS options, backup retention, SSH access method, or security documentation. Recheck it after a major application release, infrastructure change, suspected incident, or change in who administers the site. Keep a dated record with the reviewer, findings, corrective actions, and next review date.
Next action: create a copy of this checklist, list every hosting and domain account, enable MFA, remove one unnecessary access path, and schedule a test restore. Those four steps establish a practical baseline that you can improve during the next security review.