168 CloudFlare Nodes in 85 of totally 130 Countries

Visit this page regularly as it is updated daily. Last data update: 31-08-2026 09:00:49 UTC

Africa

Asia

Europe

Bratislava: 1 Node with 12 IP Addresses

CF Code	City	Country	CF IP-Country	Node IP
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.134
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.135
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.138
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.139
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.149
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.150
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.151
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.164
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.168
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.169
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.170
BTS	Bratislava	Slovakia (SK)	Slovakia (SK)	172.70.225.171

Parse Time: 15.014 ms

WP-Rush: WordPress Performance by Prevention

Why TTFB and Lighthouse Scores Don’t Measure the Same Problem

This post aims to unpack a common WordPress performance question using a neutral, first-principles lens. It is not a product recommendation; it is an attempt to separate layers, costs, and trade-offs so the discussion can be more precise.

Two metrics, two worlds

TTFB (Time to First Byte) is primarily a server and network metric: how long it takes before the first response byte arrives.
Lighthouse scores aggregate multiple lab metrics about how content renders and becomes interactive, heavily influenced by front-end workload.
When someone reports a big improvement, it helps to ask: did they reduce CPU work, reduce I/O, reduce network transfer, or simply change what was measured?
A practical way to keep the debate grounded is to define what you mean by “faster.” For some teams, the business metric is conversion; for others, it is crawl efficiency or editorial workflow. Different goals favor different interventions.

Why confusion happens

Both are discussed under the umbrella term ‘performance.’ People assume improvements in one should imply improvements in the other.
In practice, you can improve Lighthouse while worsening TTFB, or improve TTFB while leaving Lighthouse almost unchanged.
When someone reports a big improvement, it helps to ask: did they reduce CPU work, reduce I/O, reduce network transfer, or simply change what was measured?
If you are comparing approaches, control what you can: same origin server state, same test location, same cache state, and multiple samples. Otherwise, you are mostly measuring randomness.

How to think about the main document

The main document is the HTML response that starts the browser’s work. Its generation time matters for TTFB.
Its contents—scripts, styles, markup weight—shape Lighthouse metrics. The server can be fast while the payload is heavy, or slow while the payload is light.
Try to avoid all-in narratives. Most sites need a combination of techniques; the useful part is knowing which technique addresses which bottleneck.
When someone reports a big improvement, it helps to ask: did they reduce CPU work, reduce I/O, reduce network transfer, or simply change what was measured?

Trade-offs in common optimizations

HTML minification can reduce bytes but add CPU time on the server, potentially increasing TTFB.
Inlining critical CSS can speed first paint but can increase HTML size and slow down backend rendering, depending on implementation.
Try to avoid all-in narratives. Most sites need a combination of techniques; the useful part is knowing which technique addresses which bottleneck.
In WordPress specifically, small design choices—autoloaded options, hook priority, filesystem checks—can have outsized impact because they occur on nearly every request.

Why ‘score chasing’ becomes a trap

Scores reward what they measure. If the score ignores backend cost, it can encourage adding backend work to improve front-end timing.
Again, this is not an accusation. It is a predictable outcome of optimizing against a proxy metric.
Try to avoid all-in narratives. Most sites need a combination of techniques; the useful part is knowing which technique addresses which bottleneck.
When someone reports a big improvement, it helps to ask: did they reduce CPU work, reduce I/O, reduce network transfer, or simply change what was measured?

A neutral reporting approach

Report at least three numbers: server generation time (or TTFB under controlled network), HTML transfer size, and a rendering metric such as LCP.
If you want to compare architectures, keep the test location consistent and run multiple samples to avoid outliers dominating the conclusion.
Try to avoid all-in narratives. Most sites need a combination of techniques; the useful part is knowing which technique addresses which bottleneck.
In WordPress specifically, small design choices—autoloaded options, hook priority, filesystem checks—can have outsized impact because they occur on nearly every request.

Where prevention fits

If you can avoid executing unnecessary backend work for a request, TTFB can improve without manipulating the HTML.
If you can reduce front-end workload structurally (fewer assets, less script), Lighthouse can improve without post-processing.
In WordPress specifically, small design choices—autoloaded options, hook priority, filesystem checks—can have outsized impact because they occur on nearly every request.
If you are comparing approaches, control what you can: same origin server state, same test location, same cache state, and multiple samples. Otherwise, you are mostly measuring randomness.

Where this shows up in practice

In day-to-day troubleshooting, the fastest path to clarity is often to pick one representative URL and follow it end to end: request in, code executed, data fetched, HTML produced, assets requested, pixels painted.
If the conversation stays at the level of plugin brands and scores, it is easy to miss the actual bottleneck. A single trace or profile can often replace pages of speculation.
If you are comparing approaches, control what you can: same origin server state, same test location, same cache state, and multiple samples. Otherwise, you are mostly measuring randomness.
If you are comparing approaches, control what you can: same origin server state, same test location, same cache state, and multiple samples. Otherwise, you are mostly measuring randomness.
Neutral framing does not mean indecision. It means you can make a decision based on observed constraints rather than inherited slogans.

Discussion prompts

If you reply, consider sharing measurements and constraints. Clear context tends to produce better answers than generic declarations.
Which of your recent changes improved a score but increased backend work?
Do you have a shared definition of ‘fast’ across stakeholders: developers, SEO, and business owners?

Key takeaways

Separate backend generation time from frontend rendering time; they respond to different interventions.
Ask whether a change reduces work, shifts work, or adds work after the fact.
Treat caching as a powerful tool, but not a substitute for understanding miss-path cost.
Consider request classification as a neutral framing for deciding what must execute.