Reactive Linux Support vs Preventive Maintenance

AV Services · avservices.in · Linux Infrastructure, Mumbai · Since 1999

Most businesses run on reactive support without knowing it. The server works. Nobody calls anyone. Then it does not work — and everything happens at once. This page is about what that cycle costs versus what preventing it costs.

What reactive support actually looks like

Reactive support is not a decision most businesses make consciously. It is the default. Nobody sets up a server and thinks “we will only call someone when it breaks.” It just ends up that way. The server runs, months pass, nothing breaks visibly, so nobody does anything.

Meanwhile, the disk fills slowly. Packages accumulate unpatched CVEs. A cron job that was running correctly breaks quietly after a path change. The backup job runs every night but the destination ran out of space six weeks ago — and the job completes with an error nobody reads.

None of these announce themselves. They accumulate silently. Then one morning the ERP does not start, or the database refuses connections, or the server is simply unreachable. The incident that follows feels sudden. It has been building for months.

What reactive support costs when it activates

Emergency Linux response is priced differently from retainer work. It should be — the engineer drops everything, responds immediately, works under pressure without prior knowledge of the environment.

At AV Services, the emergency rate for non-retainer clients is Rs.15,000 for the first 3 hours onsite in Mumbai. Data recovery from a formatted or corrupted volume runs Rs.50,000-75,000 depending on complexity. Full ERP restoration across multiple visits is Rs.40,000-60,000.

That is one incident. One bad morning. One corrupted volume or one full disk or one failed update on a production server.

And those are just the direct costs. Add the indirect ones: production stopped, staff idle, orders delayed, clients calling, the operations team managing the crisis instead of running operations. A 6-hour ERP outage at a Mumbai manufacturing firm with 50 employees costs far more than the recovery bill.

What preventive maintenance actually catches

The value of prevention is invisible by design. Nothing dramatic happens. The disk alert fires at 80%, gets addressed, never reaches 100%. The backup destination gets cleared before the job starts failing. The kernel gets patched before the CVE gets exploited.

These are not hypotheticals. They are the standard monthly findings from any managed Linux server. In 25 years of infrastructure work in Mumbai, the pattern is consistent: servers that are actively maintained produce incidents at a fraction of the rate of servers that are not.

The AV Services infrastructure failures we have prevented page documents real examples — specific incidents that were caught and resolved before they became outages. Every one of them would have been an emergency without proactive monitoring.

The cost comparison

Factor	Reactive Support	Preventive Retainer
Monthly cost	Rs.0 — until it is not	Rs.15,000-50,000
Cost per incident	Rs.15,000-75,000+ per event	Included in retainer
Incident frequency	Unpredictable — often clusters	Significantly reduced
Response time	Finding someone — hours	10 min remote · 4 hrs onsite
Engineer knows your server	No — starts cold every time	Yes — documented history
Backup verification	Assumed — rarely tested	Tested monthly
Security patching	When someone remembers	Monthly — documented
After-hours availability	Find someone — good luck	24/7 incl. weekends

The maths most businesses do not run

A business on reactive support pays Rs.0 per month until an incident hits. One data recovery — Rs.60,000. One ERP restoration — Rs.50,000. Total for one bad year: Rs.1,10,000 in direct recovery costs, plus the production loss, plus the stress, plus the discovery that the last clean backup was 6 weeks ago.

The Professional retainer is Rs.30,000 per month — Rs.3,60,000 per year. It sounds like more. It usually is not, once incidents are counted honestly. And the incidents on a well-maintained server are a fraction of those on an unmanaged one.

The businesses that switch to retainer after an incident always say the same thing: they wish they had done it before the incident, not because of it.

When reactive support is acceptable

Reactive support works when downtime is genuinely tolerable. A development server used only during office hours. A staging environment. A personal project server. Infrastructure where the cost of a 24-hour outage is minor inconvenience, not business loss.

For production ERP, manufacturing operations, trading infrastructure, or any system where downtime means money stopping — reactive support is not a cost saving. It is a deferred cost with an unknown due date.