Illustrative scenario: early-stage B2B SaaS company with a small in-house team
Bringing a SaaS platform's dependencies and servers up to date
A B2B SaaS company's product is growing, but its framework, dependencies and servers have fallen behind while the team focused on features. This scenario shows how SupportSafe would bring the platform up to date in controlled, reversible steps.
This is an illustrative scenario that shows how we would approach a typical engagement. It does not describe a specific client, and outcomes describe what the work is designed to achieve and how success would be measured.
The challenge
In this scenario, a B2B SaaS company runs a web application built on a popular PHP framework with a JavaScript front end, a PostgreSQL database and a background job queue. It is hosted on a small number of Linux virtual servers that were set up by a founding engineer. The in-house team of four developers is focused on product features demanded by customers, and infrastructure work has repeatedly been pushed down the backlog.
What the business is experiencing
- Security questionnaires it struggles to answer. Larger prospects ask about patching, backups and access control, and the honest answers are uncomfortable.
- Dependency drift. The framework version is out of active support, and automated tools flag dozens of outdated packages, some with known vulnerabilities.
- Fragile servers. The operating system is approaching the end of its standard support period, kernel and package updates have not been applied for months, and nobody is confident about rebooting the servers.
- Slow background jobs. Report generation and email notifications back up during busy periods, and customers occasionally notice.
What the technical review is likely to find
- A framework major version that no longer receives security fixes, blocking upgrades of several other libraries.
- Front-end build tooling pinned to old versions because a newer release changed configuration.
- Servers configured by hand with no record of what was installed or why, and SSH access via shared keys.
- Database backups taken nightly with a dump script, but no record of a restore ever being tested, and no defined recovery point objective (RPO).
- Missing database indexes on frequently filtered tables, and long-running queries during report generation.
- Monitoring limited to an external uptime check, with no visibility of disk space, memory, queue length or error rates.
Why it matters
For a SaaS company, the platform is the product. Unpatched dependencies and undocumented servers are both a security risk and a sales obstacle, and an untested backup puts every customer's data at risk. The team needs to modernise the platform without pausing feature work and without a risky, all-at-once migration.
Our approach
SupportSafe would work alongside the in-house developers, taking ownership of infrastructure and maintenance while the product team keeps shipping features. Every change is made in small, reversible steps.
Assess
- Review the codebase dependencies, framework version, build pipeline, servers, database and backup arrangements.
- Produce a dependency upgrade map showing which packages block others and which have known vulnerabilities.
- Agree target RPO and RTO values with the business, based on what customers would reasonably expect.
Onboard
- Replace shared SSH keys with individual keys, restrict access by role and record all access in a credential vault.
- Add server and application monitoring: CPU, memory, disk, database connections, queue length, error rates and SSL expiry.
- Run a full restore of the production database to an isolated server, document the steps and time taken, and compare the result with the agreed RTO.
Maintain: dependencies
- Upgrade dependencies in small batches through the existing pull-request workflow, so the in-house team reviews every change.
- Plan the framework upgrade one major version at a time, using the automated test suite and adding tests around critical paths where coverage is thin.
- Update front-end build tooling and address deprecation warnings before they become breaking changes.
Maintain: servers and database
- Capture current server configuration as code so it can be rebuilt consistently.
- Build replacement servers on a supported operating system release, test the application on them, then switch traffic over with a rollback route available.
- Apply a regular patching schedule with planned maintenance windows communicated to customers where needed.
- Add missing indexes, tune slow queries and review queue worker configuration.
- Set up encrypted, off-site database backups with point-in-time recovery where the hosting allows, and schedule regular restore tests.
Work with the in-house team
- Agree a shared change calendar so infrastructure work never collides with product releases.
- Document every server, service and runbook in the team's own knowledge base, so knowledge no longer sits with one engineer.
- Hold short fortnightly check-ins during the upgrade phase, then move to monthly reviews once the platform is stable.
Report and improve
- Monthly reports covering patches applied, dependency status, backup verification and incidents.
- A quarterly roadmap covering remaining technical debt, capacity planning and readiness for customer security reviews.
Intended outcomes and how they're measured
This engagement is designed to bring the platform onto supported, documented foundations and to give the company credible answers to customer security questions. The outcomes below describe what the work aims to achieve and how success would be measured; they are not results from a real company.
Supported foundations
- Aim: every major component is on a version that still receives security fixes.
- Success measured by: framework, language runtime, operating system and database on supported releases; dependency scanning showing no unaddressed critical or high-severity vulnerabilities, with any exceptions documented and risk-accepted.
Proven recoverability
- Aim: the company knows, rather than assumes, that it can recover customer data.
- Success measured by: a restore test completed on a regular schedule, with the measured recovery time within the agreed RTO and data loss within the agreed RPO; backups stored off-site and encrypted.
Reduced operational risk
- Aim: problems are detected early and servers can be rebuilt consistently.
- Success measured by: monitoring alerts in place for resource usage, queue length and error rates; server configuration held as code; a documented rebuild procedure tested at least once.
Better performance for background work
- Aim: reports and notifications are processed promptly during busy periods.
- Success measured by: queue length and job duration tracked over time against the baseline captured during assessment; slow-query logs reviewed monthly.
Sales and compliance readiness
- Aim: the team can answer security questionnaires accurately.
- Success measured by: documented processes for patching, access control, backups and incident response that the company can reference in customer reviews. SupportSafe supports the technical controls behind frameworks such as GDPR or SOC 2 but does not provide legal or compliance certification advice.
No engineering programme can promise that a platform will never fail. The aim is to make failures less likely, easier to detect and quicker to recover from.
Services used in this scenario
SaaS Maintenance
Release discipline, dependency upgrades, background job health and incident response for multi-tenant SaaS platforms, alongside your product team.
Learn moreLinux Server Maintenance
Kernel and package patching, SSH hardening, fail2ban, log rotation and systemd health for Ubuntu, Debian, AlmaLinux and Rocky servers.
Learn moreDatabase Maintenance
Verified backups, index health, version upgrades and replication checks for MySQL, PostgreSQL, MariaDB and MongoDB databases behind your applications.
Learn moreSecurity Patch Management
Tracked, tested and staged security updates for operating systems, CMS platforms, plugins and libraries, with CVE monitoring and rollback plans.
Learn moreFacing a similar situation?
Tell us about it. We'll assess your setup and explain how we'd approach it – before you commit to anything.