Seven Quiet Signals Your Hosting Is Falling Apart (Before It Actually Does)

Here's the thing nobody tells you when you sign up for a hosting plan: the catastrophic outage you're afraid of almost never shows up unannounced. It sends scouts first. Slow page loads at 2 PM on a Tuesday. A database query that used to run in 80 milliseconds now taking 400. A DNS lookup that hesitates just long enough to make you wonder.

Most founders see these things and assume it's their code, their browser, their internet connection — anything but the hosting. By the time the actual meltdown happens, the warning signs have been flashing for weeks.

This guide is about reading those signs before the emergency arrives.

1. Your Slow Query Log Is Talking. Are You Listening?

If you're running any kind of database-backed site — WordPress, a custom app, an e-commerce store — your database has a slow query log. Most cheap hosting dashboards bury it or don't surface it at all, but if you have SSH access or a tool like phpMyAdmin, you can find it.

A slow query log captures any database operation that takes longer than a defined threshold (usually one second). One or two entries? Normal. A log that's growing by hundreds of entries per day? That's your hosting environment degrading under load.

What you're looking for: repeated queries on the same tables, full table scans where indexes should exist, and queries that suddenly got slower without any code changes on your end. That last one is the kicker — if your code didn't change but your query times did, the problem is almost certainly the shared environment your host crammed you into.

Quick action: Enable slow query logging if it isn't on. Check it weekly. Set a calendar reminder if you have to.

2. CPU Spikes That Don't Match Your Traffic

Log into your hosting control panel and pull up your resource usage graphs. Most decent hosts — even budget ones — will show you CPU utilization over time. What you're looking for is spikes that don't correspond to actual traffic surges.

If your analytics show 50 visitors between 3 and 4 AM but your CPU graph shows a 90% utilization spike during that same window, something is happening that you didn't ask for. It could be a rogue plugin running a scheduled task, a bot hammering your login page, or — and this one's sneaky — your hosting environment running maintenance jobs that compete with your site's resources.

On shared hosting, you're also potentially inheriting the resource chaos of your neighbors. That CPU spike might not even be yours. Either way, unexplained spikes are a diagnostic red flag worth investigating before they become a consistent performance problem.

3. DNS Propagation Delays That Linger Too Long

When you make a DNS change — updating an A record, switching nameservers, pointing a subdomain somewhere new — propagation is supposed to take anywhere from a few minutes to 48 hours. That's normal. What's not normal is changes that seem to propagate in some regions but not others, or changes that appear to propagate and then revert.

Use a tool like whatsmydns.net to check how your DNS records are resolving across different global locations. If you're seeing inconsistency that stretches beyond 48 hours, or if records are resolving differently across US-based servers that should all be pulling the same data, your hosting provider's DNS infrastructure has a reliability problem.

This matters more than most founders realize. Inconsistent DNS means some of your visitors are landing on a working site while others are hitting a dead end — and they're not going to email you about it. They're just going to leave.

4. Time-to-First-Byte Creep

Time-to-First-Byte (TTFB) is how long it takes your server to start responding after a browser sends a request. A healthy TTFB for a shared hosting environment is under 600 milliseconds. Under 200ms is excellent. Over a second is a problem.

The sneaky part? TTFB creep happens gradually. Your site might have launched with a 300ms TTFB and slowly crept to 900ms over six months as your host's shared server got more crowded. You didn't notice because the change was incremental.

Run a TTFB test today using Google's PageSpeed Insights or a tool like GTmetrix. Then run it again in 30 days. If the number is trending upward without any significant changes to your site, your hosting environment is getting worse — not your code.

5. PHP Memory Limit Errors Showing Up in Your Logs

WordPress and most PHP-based applications will throw memory limit errors when the server can't allocate enough RAM to complete an operation. These errors often get swallowed silently — your site appears to work, but background processes fail, images don't resize correctly, or plugin updates bail out halfway through.

Check your PHP error logs. If you're seeing Fatal error: Allowed memory size exhausted appearing regularly, your host has set a memory ceiling that your site is brushing up against. On many budget plans, the default PHP memory limit is 64MB or 128MB — fine for a static brochure site, genuinely insufficient for anything dynamic.

The fix is sometimes as simple as adding php_value memory_limit 256M to your .htaccess file, but whether that actually works depends on whether your host allows it. If they don't, that's a constraint worth knowing about now.

6. SSL Certificate Renewal Warnings You Didn't Get

Let's Encrypt certificates expire every 90 days and are supposed to auto-renew. On a well-configured hosting environment, this happens invisibly. On a poorly maintained one, auto-renewal quietly fails — and the first you hear about it is when your visitors start seeing browser security warnings.

Set a manual calendar reminder 30 days before your SSL cert expires. You can check your expiration date by clicking the padlock icon in Chrome and viewing the certificate details. If your cert is within 30 days of expiry and you haven't gotten any renewal confirmation from your host, don't wait — reach out to support now, not the day it expires.

7. Resource Creep in Disk and Bandwidth Usage

Pull up your hosting account's usage dashboard and look at your disk space and bandwidth consumption over the past three months. Is it growing at a rate that makes sense for your actual content? Or is it growing faster than you'd expect?

Unexplained disk growth can indicate log files that aren't being rotated, temporary files that aren't being cleaned up, or — in worse cases — your server being used for something you didn't authorize. Bandwidth spikes can mean you're getting scraped, bot-hammered, or that your media files are being hotlinked from elsewhere.

Build Your 15-Minute Monthly Audit

None of these checks take long individually. String them together into a monthly 15-minute routine:

• Check slow query log for new entries
• Review CPU usage graphs for unexplained spikes
• Run a TTFB test and compare to last month
• Verify DNS consistency with whatsmydns.net
• Check PHP error logs for memory warnings
• Confirm SSL cert expiration date
• Review disk and bandwidth usage trends

The goal isn't to become a systems administrator. The goal is to catch the decay before it becomes a disaster — because on a free or budget hosting plan, the margin between "running fine" and "down on a Friday afternoon" is thinner than the sales page will ever admit.