It’s not Cloudflare: it’s LALIGA. Why your website goes wrong when Barça plays.
What’s going on with LALIGA and why it affects you even if you don’t watch soccer online
At Damos en el Blanco we prefer to explain things without smoke and mirrors. In the last few weeks we have seen intermittent crashes and capricious accesses in some clients’ websites with traffic from Spain. And, although there was a punctual Cloudflare incident on August 21, what has impacted us the most is not explained by that. What we see -and media and associations confirm- is a blocking of IP ranges associated with LALIGA that is affecting thousands of legitimate sites as collateral damage.
What we have experienced (in simple)
- Peaks of users who could not enter or entered at times.
- Falsified campaign metrics (traffic that “disappears” due to blocking, not low demand).
- Support overflowed: “is my website broken?”. No, they are blocking it through no fault of your own.
The worrying thing is that we are not talking about websites that broadcast soccer: we are seeing the effect on legal businesses, institutions and cultural sites. Internet users’ associations have filed formal complaints about these massive and indiscriminate blockages with the Ombudsman, precisely because of the damage to pages that are totally unrelated to piracy.
“So it’s Cloudflare’s fault?” no.
Separating issues is key:
- Yes there was an isolated Cloudflare incident on August 21 for congestion with AWS us-east-1. It was global, had a clear beginning and end, and Cloudflare published its post-mortem. It does not explain the recurring crashes in Spain that we describe here.
- The widespread in Spain coincides with the start of the 2025/26 season and with blocking measures that take away websites that have nothing to do with pirate soccer. Several media have documented it; LALIGA publicly denies it, but the damage is there and affects innocent third parties.
Why it’s so unfair (and dangerous)
Because it cuts to the chase: if an IP range “smells” of something it doesn’t like, everything that travels that way is blocked, including stores, media and organizations. The result:
- Lost sales in e-commerce.
- Reputation affected (“your website is going bad”).
- Contaminated data for decision making.
This cannot be normalized: fighting piracy is legitimate, breaking the Internet along the way is not. That is why there are formal complaints asking for intervention from the authorities and a proportionate framework.
What we are doing (our customer plan)
- Mitigating the damage today
- Alternate routes and multi-CDN when applicable, to bypass blocked segments.
- Degraded modes (essential content usable even with partial blocking).
- Status pages and warnings to avoid the “it’s broken” feeling.
- Observe and test
- Monitoring by operator/country to detect blocking patterns.
- Pre-peak load tests (sports, launches, campaigns).
- Defending your interests
- Record of evidence (dates, IPs, traceroutes, captures) to claim.
- Coordination with other affected parties and associations that have already made representations to the Ombudsman.
If it is affecting you
- Write to us. We activate a contingency plan in your hosting/CDN.
- We help you to document it for possible claims.
- We adjust your analytics to separate “blocking crashes” from actual user behavior.
Contact: [email protected] – 931 89 00 35
For those who want to go deeper (readings)
- Startup 2025/26 and IP blocking with impact to legitimate websites. Xataka
- Complaint by AI, UAI and FED to the Ombudsman for indiscriminate blockades. Cinco Días
- Coverage on collateral damage and public claims. THE BULLETIN Confilegal
- Cloudflare post-mortem (08/21/2025) – separate incident from local blocking. The Cloudflare Blog
- LALIGA’s position (denies massive blockades). www.20minutos.es – Last News
A year on: what we have learned absorbing spikes
That episode was about someone else’s blocking — something that lands on you and you cannot control. But there is a second half to the story that you do control: how your site is built when the hit arrives. We see that one every time a client gets real traffic.
The most extreme case we have served is eclipsi.info, on 12 August 2026. Look at the shape of the curve: it climbs all afternoon and explodes at eclipse hour.
See the data as a table
| Hour | Requests |
|---|---|
| 02:00 | 9,784 |
| 03:00 | 3,776 |
| 04:00 | 3,970 |
| 05:00 | 4,635 |
| 06:00 | 11,918 |
| 07:00 | 34,540 |
| 08:00 | 77,449 |
| 09:00 | 106,766 |
| 10:00 | 121,906 |
| 11:00 | 139,783 |
| 12:00 | 144,016 |
| 13:00 | 154,697 |
| 14:00 | 178,118 |
| 15:00 | 205,904 |
| 16:00 | 242,815 |
| 17:00 | 285,719 |
| 18:00 | 420,745 |
| 19:00 | 570,865 |
| 20:00 | 171,640 |
| 21:00 | 29,424 |
| 22:00 | 23,994 |
| 23:00 | 16,713 |
| 00:00 · 13 | 10,273 |
| 01:00 · 13 | 5,692 |
| 02:00 · 13 | 6,261 |
570,865 requests in one hour, about 158 per second, inside a total of 8.3 million over ten days. The result: 666 server errors, or 0.008%. 94.6% of the data came from cache and the origin barely noticed.
That is not an expensive server. It is three decisions taken beforehand: that the site never computes anything server-side, that everything is cacheable behind a CDN, and that cache-busting is tested before the big day. The same decisions that, when an IP block like LALIGA’s arrives, leave you room to move the affected piece instead of watching everything fall over.
And in practice
We build and run that part: servers, CDN, caching, backups, certificates and monitoring. It is managed hosting, not a control panel where you fend for yourself. If you have a date circled in red — a launch, a campaign, a big day — that is exactly when to look at it, not the day after.
Where to go next: the full numbers from that day · the eclipsi.info case · what our hosting covers