31 mins read

Famous Attacks on Web Hosting: Key Lessons Learned and How to Stay Protected

The famous attacks on web hosting companies share one pattern. They rarely began with anything exotic. A stolen password, a missing second login step, an unpatched system or a backup sitting next to the data it was supposed to protect did the damage. The lessons are practical, and most of them apply to a single site owner as much as to a large provider.

We have run hosting infrastructure for over 20 years, so we read these incident reports the way a mechanic reads a crash report. This guide covers five verified cases: Code Spaces, OVH, GoDaddy, the 2019 hosting ransomware wave and the OVHcloud Strasbourg fire. For each one we explain what happened, who it hurt, which weakness mattered and what you can do about it.

Two cautions before we start. The Strasbourg fire was not a cyberattack, and we include it because it teaches the same backup lesson. And responsibility is shared. A secure provider helps a great deal, but your own account security, independent backups and tested recovery plan still decide how bad a bad day gets.

What Are the Most Common Cyberattacks Targeting Web Hosting Companies?

The most common are DDoS attacks, ransomware, data breaches through stolen credentials, exploited server vulnerabilities and supply chain attacks. Most of them start with something ordinary, like a reused password or an unpatched system. Hosting companies are attractive targets because one successful break in can reach thousands of customers at once.

DDoS attacks that overwhelm hosting infrastructure

A distributed denial of service attack, or DDoS, floods a target with so much traffic that real visitors cannot get through. The goal is usually to knock a site offline, to extort money or to distract the defenders while something else happens.

That last use is the one to remember. The Code Spaces attack in 2014 began with a DDoS, and the company later learned that the flood was only part of the picture. Someone had also reached its cloud control panel and was leaving messages.

A CDN in front of your site can soak up a large share of junk traffic before it reaches your server. We offer CloudFlare CDN for that reason. A CDN reduces the pressure. It does not fix a stolen password or a weak backup plan, which is where the bigger damage came from in every case below.

Ransomware attacks that encrypt servers and customer data

Ransomware encrypts files and demands payment for the key. Against a hosting provider, it can lock many customers’ sites and databases in one stroke, because they all sit on the same infrastructure.

A threat memo from CERT EU, the computer emergency team for European Union institutions, reported that more than ten web provider companies had suffered targeted ransomware incidents in 2019 alone. It also noted that the largest known ransom payment came from a hosting provider, a South Korean company called Nayana, which reportedly paid about $1.14 million in 2017. Criminals read news like that. A CERT EU analysis warned that more attacks were likely to follow.

Data breaches involving customer accounts and credentials

Customer data breaches are the quietest kind. Nothing goes offline. An attacker simply copies names, email addresses, hosting passwords or private keys, and uses them later.

OVH in 2013 and GoDaddy in 2019 and 2021 both fit this pattern, and we cover them below. What those leaks enable is the real danger. A stolen hosting login lets an attacker change your site, steal your visitors’ data or send spam from your domain. Email is often the first weak point, so clean inbound mail matters. We offer MailChannels email, which is built to keep spam and phishing away from mailboxes.

If your credentials leak, change the password everywhere you reused it, not just on the affected service. Attackers test stolen logins against other sites within hours, and a reused password turns one provider’s breach into several of yours.

Server vulnerabilities that allow unauthorized access

Server vulnerabilities are flaws in operating systems, control panels, web servers and applications that let an attacker in without a password. Remote code execution, where the attacker runs commands on the server, is the worst kind. A web shell, a small malicious script uploaded to a site, often gives them a permanent foothold afterward.

One report on GoDaddy’s 2019 breach said the intruder likely used an unpatched vulnerability in a customer managed hosting environment. The point is not that one company failed. The point is that a single unpatched system, in an environment nobody was watching closely, was enough.

Control panels deserve special attention because they run with high privileges and face the internet. A flaw there is worth more to an attacker than a flaw in a single plugin. Apply panel and operating system updates as soon as you can test them, and treat any security notice from your panel vendor as urgent.

Supply chain attacks targeting hosting platforms and connected services

In a supply chain attack, the criminal goes after a tool or vendor that many victims trust, instead of attacking each victim. Control panels, management software and service providers are all candidates. The 2021 Kaseya incident showed how an attack on one management product could reach the customers of the managed service providers that used it.

Hosting has its own version. The FTC’s account of a December 2022 GoDaddy incident described a threat actor who compromised a cPanel service and redirected some visitors of customer sites to malicious pages. The sites themselves had not been hacked directly. The platform underneath them had.

For a site owner, this is a hard risk to control. You can choose vendors carefully, keep their software updated and limit what each one can reach. You cannot audit them all.

What Are Some Famous Cyberattacks on Web Hosting Providers?

Five cases teach the most: Code Spaces in 2014, OVH in 2013, GoDaddy from 2019 onward, the 2019 ransomware attacks on A2 Hosting, iNSYNQ and SmarterASP.NET, and the OVHcloud Strasbourg fire in 2021. The first four were attacks. The fire was a physical disaster, and it belongs on this list because it taught the same lesson about backups.

Code Spaces (2014): How compromised cloud access destroyed a hosting business

Code Spaces hosted source code repositories for developers on Amazon Web Services. On June 17, 2014, it was hit by a DDoS attack, and the company then found messages from someone who had gained access to its EC2 control panel. The messages led to an extortion demand.

The company tried to take back control by changing passwords. According to its own statement, the attacker had already created extra logins and then began deleting things from the panel. By the end, snapshots, storage buckets, machine images and some servers were gone, along with backups the company had described as off site. Code Spaces could not recover and shut down.

The weakness was plain. The company’s own account suggested a compromised username and password, and it had not used two factor authentication on the panel. The backups lived inside the same console as the live systems, so whoever controlled the console controlled everything. As Help Net Security noted, customers pointed out that the backups were reachable from that one panel.

The lesson for your own setup: protect the account that controls everything with MFA, and keep at least one backup copy under different credentials that the main login cannot delete.

OVH (2013): How compromised employee credentials exposed customer information

In July 2013, the French hosting provider OVH told customers that its internal network had been breached. The attacker first got into the email account of a system administrator, used that access to reach another employee’s internal VPN, and then compromised the access of an administrator who handled the back office.

The result was a copy of OVH’s European customer database. It held names, addresses, phone numbers and passwords that were hashed and salted, and OVH said no credit card data was stored there or viewed. The attacker also reached an installation server in Canada. OVH asked customers to change their passwords and filed a criminal complaint.

The weakness was a chain of trust. OVH’s own description said staff access relied on two checks, a password and the source network. Once one email account fell, the chain followed. OVH responded by regenerating employee passwords, moving critical access behind a new VPN, restricting email access and adding a hardware security token as a third check. It also published a detailed account of what happened, which The Register reported and which remains one of the more open disclosures this industry has seen.

The lesson: treat email as a master key, because password resets flow through it. Protect it with MFA, and never let one compromised account hold the keys to the next.

GoDaddy (2019 to 2021): Lessons from repeated hosting environment compromises

GoDaddy’s story is several incidents, and they should be kept apart. In October 2019, an intruder began connecting to about 28,000 customers’ hosting accounts over SSH. GoDaddy disclosed it in 2020 and the access had gone unnoticed for roughly six months, according to later reporting.

In November 2021, an attacker used a compromised password to reach GoDaddy’s Managed WordPress hosting environment. The data of about 1.2 million customers was exposed, including email addresses, WordPress admin passwords, sFTP and database credentials, and SSL private keys for some customers, as BleepingComputer reported. The FTC later tied these incidents, and another one in December 2022, to the same period of weak security.

The FTC alleged several weaknesses. It cited a failure to keep an inventory of assets, inadequate separation of the shared hosting environment from less secure networks, and unsecured connections to services such as APIs. A summary of the complaint described the 2019 breach as one that started in a customer managed environment and spread into shared hosting because of poor segmentation.

The FTC’s final order, finalized in 2025, requires GoDaddy to run a formal security program, bars it from misleading customers about security, and calls for multi factor authentication for staff and customers. Notice what that implies. A very large company still needed regulators to push it on the basics.

SmarterASP.NET, iNSYNQ and A2 Hosting (2019): The impact of ransomware on hosting providers

Three providers were hit by ransomware in one year. A2 Hosting suffered an attack in May 2019, in which a version of GlobeImposter ransomware encrypted servers in Asia and North America. iNSYNQ, a cloud provider that hosts accounting software, was hit by MegaCortex in July. CERT EU reported that both companies took weeks to restore customer data.

SmarterASP.NET followed in November. The company hosted more than 440,000 customers, and its own notice told customers that their hosting accounts had been encrypted. SecurityWeek reported that it was recovering customer data over the following days. The attack reached backend databases as well as public web servers, so customers who used the service as an application backend could not simply move elsewhere.

Public reports do not make clear how the attackers got in, and we will not guess. What the three cases show is the customer side. Websites, databases and sometimes the provider’s own support channels went down together. Anyone who relied only on the provider’s backups waited with everyone else.

The lesson: a provider outage becomes your outage. Keep your own recent copy of your site and database somewhere the provider’s systems cannot touch.

OVHcloud Strasbourg (2021): What a major data center fire revealed about infrastructure resilience

This was not a cyberattack. In the early hours of March 10, 2021, a fire destroyed the SBG2 data center building at OVHcloud’s Strasbourg campus and badly damaged SBG1, according to reporting on the investigation. Sites hosted there went offline, and many customers lost data permanently.

The reason is the backup lesson in its purest form. Many affected customers had not bought backups, and some who had paid for them found the copies were stored in the same building as the live servers. A French court later ordered OVHcloud to pay a combined €250,000 to two customers whose paid backups were lost in the same fire. OVHcloud is appealing, and courts dismissed claims from other customers who had no backup service at all.

OVH’s founder said afterward that the company would provide backups by default across its data centers, at no extra cost. Whatever your view of that decision, the practical lesson holds. A backup in the same building as the original protects you from deleted files and nothing else. Disaster recovery means a copy in another location.

What Security Weaknesses Made These Hosting Attacks Possible?

The same five weaknesses appear again and again: weak or stolen administrative credentials, too many permissions, flat networks where one breach reaches everything, unpatched software and backups that sit within reach of the same accounts as the live data. None of them is a secret, and every one can be reduced.

Weak authentication and compromised administrative credentials

Code Spaces had no second login step on the panel that controlled its whole business. OVH’s staff access relied on a password and a network check, and one hijacked email account was enough to start a chain. GoDaddy’s 2021 incident began with a compromised password.

Passwords fail in predictable ways. People reuse them, phishing steals them and breaches leak them. MFA does not stop all of that, but it turns a stolen password from a complete failure into a partial one.

Apply it first to the accounts with the most reach. Your hosting control panel, your domain registrar, your email and your cloud console.

Excessive permissions and insufficient access restrictions

At Code Spaces, whoever held the panel login could delete the live servers and the backups with it. In the OVH chain, the attacker moved from an ordinary administrator’s email to an administrator who handled the back office. Each step gained more power than the last, and nothing in the way said stop.

Least privilege means every account gets only the permissions its job needs. A developer who deploys code does not need to delete backups. The account that creates backups should not be able to erase them.

This costs some convenience. It means more accounts and more switching. We think that trade is cheap compared with losing the lot.

Inadequate network segmentation and server isolation

Segmentation divides a network into zones, so a breach in one zone cannot freely reach the others. The FTC alleged that GoDaddy did not adequately separate its shared hosting environment from less secure networks, and that this let a 2019 breach spread. In the 2019 ransomware cases, encryption spread across many servers, including regions as far apart as Asia and North America at A2 Hosting.

On shared web hosting, many customers share one server, and the provider is responsible for isolating accounts from each other. Isolation is a security boundary, and it is only as good as its configuration. Ask your provider how it separates accounts and what happens if one is compromised.

Vulnerable software and delayed security patching

Unpatched software is the oldest weakness on this list. GoDaddy’s 2019 intrusion was reported to involve an unpatched system in a customer managed environment. The FTC also listed missing asset inventory among its findings, and that matters because you cannot patch what you do not know you have.

The same logic applies to a small site. A forgotten plugin, an old test install or an outdated PHP version keeps working quietly until someone finds it. Patching needs a routine and a list of everything that runs on your server.

Automation helps with patching. It does not replace knowing what you run.

Backups exposed to the same systems and administrative accounts as production data

Code Spaces and the Strasbourg customers lost for the same reason. Their backups were reachable by the same login or sat in the same building as the originals. When production failed, so did the safety net.

Ransomware has the same effect in a different way. When the attacker reaches connected storage, encryption spreads to anything the infected systems can write to. Backups need distance: a different location, different credentials and ideally settings that prevent deletion for a period.

If one stolen password can erase your site and every backup of it, you have one copy with extra steps.

Ask one question about every backup you hold. If someone stole my main login tonight, could they delete this copy? If the answer is yes, it is not yet an independent backup.

What Lessons Can Website Owners and Hosting Providers Learn from These Attacks?

Five lessons come out of these cases. Account security matters as much as website security, backups must be independent, monitoring catches trouble early, a practiced incident plan cuts recovery time, and honest communication from your provider builds trust. Providers carry part of this load. Site owners carry the rest.

Why hosting account security matters as much as website security

You can run a perfectly patched WordPress site and still lose it if someone takes over your hosting account. GoDaddy’s 2019 incident involved attackers using hosting credentials to connect over SSH. An attacker with account access can read files, change code and steal data, and your site’s own defenses never get a vote.

Think of two layers. The provider secures its servers and network. You secure your logins, your site software and your data. If you run a hosting business on a reseller hosting plan, you hold both roles for your customers, so your own account security is part of the product you sell.

Why independent backups are essential for business continuity

Independent means separate in three ways: a different location, different credentials and a different failure mode. Code Spaces and the Strasbourg customers lacked at least one of those. A backup that shares a building, a login or a storage system with the original fails with it.

Business continuity is the larger goal. It asks how long you can be down and how much data you can lose, and whether your backups actually meet those limits. A backup you have never restored answers none of those questions. Test one restore soon, and write down how long it takes.

Cost is the usual objection. Storage for a copy of a typical small business site costs little next to a day offline. Compare the monthly price with what one lost week of orders or content would cost you.

How server monitoring helps identify suspicious activity

The GoDaddy intrusion that began in 2019 went unnoticed for months. That is the cost of weak monitoring. Attackers who stay quiet can learn your systems, copy data and plan their next step.

Useful monitoring watches for specific signs. Logins from unfamiliar locations, new administrator accounts, changed files, sudden spikes in outbound traffic and bursts of failed SSH attempts. Keep logs somewhere an intruder on the server cannot edit them, and make sure alerts reach a person who will act.

Monitoring is also how you prove what happened. After an incident, logs tell you which account was used, what was touched and when. Without them, you restore blind and may reinstate the same weakness. Even simple tools help if someone actually reads the output.

Why incident response planning reduces recovery time

A2 Hosting and iNSYNQ took weeks to restore customer data. Not every delay can be blamed on planning, but practice changes how quickly a team moves, who decides and what gets communicated.

Code Spaces gives a sharper lesson. It said it had a recovery plan that had been tested. The plan failed anyway, because it assumed the control panel and its backups would still exist. A good plan asks what happens if the attacker already holds your main credentials, then prepares for that.

Write the steps down. Who is called, who decides, how customers are told and where the clean backup lives. Rehearse it once a year at least.

How hosting provider transparency builds customer trust

OVH told customers exactly how the 2013 attack worked, including the weakness in its own security. That disclosure helped customers act, and it taught the industry something. Compare that with reports of incidents that went undiscovered for months, followed by regulators insisting on honest statements about security.

Transparency also shows up in small things. A status page that is updated, a clear message when something breaks and support that answers. Our end user support runs 24/7 for this reason. A fast honest answer during an incident is worth more than any marketing claim.

Ask any provider you are considering how it communicates during an outage and whether it publishes incident reports. A written history of past problems and fixes is a better sign than a spotless record that nobody can verify.

How Can You Protect Your Website and Hosting Server from Cyberattacks?

Turn on MFA everywhere it matters, patch quickly, filter and scan your traffic, isolate your sites from each other and keep encrypted off site backups that you have tested. Do those five and most of the failures in these incidents no longer apply to you.

Enabling MFA and securing SSH and control panel access

Turn on multifactor authentication for your hosting control panel, domain registrar, email and any cloud console. Use an authenticator app over text messages when you have the choice. Give every account its own password and keep them in a password manager.

For SSH, use key based logins instead of passwords, disable direct root login and restrict access to known IP addresses where you can. GoDaddy’s 2019 intrusion involved SSH access, so this is not a theoretical risk. If you run billing software, protect it as well. The admin area of a WHMCS installation controls invoices, client data and provisioning, which makes it worth a second login step and an IP restriction.

Keeping server software, CMS platforms and plugins updated

Set a weekly update routine and keep it. Update the CMS, plugins, themes, PHP and your control panel. Remove anything you no longer use, since unused code still counts as attack surface.

Keep a list of what runs on each server and site. The FTC’s finding about asset inventory shows why. A server nobody remembers is a server nobody patches. When an update might break something, test it on a staging copy first, then roll it out. Delay is the expensive option.

Do not forget the hosting layer itself. On a shared plan the provider patches the server, and you patch your site. On an unmanaged VPS or dedicated server you patch both, so put the operating system and control panel on the same weekly list as WordPress.

Using firewalls, malware scanning and intrusion detection

Each tool does a different job. A firewall controls which network traffic is allowed in. A web application firewall inspects web requests for attack patterns. A malware scanner looks for malicious files such as web shells, and intrusion detection watches for suspicious behavior on the network or server.

None of them is complete alone. A CDN such as CloudFlare helps absorb DDoS traffic and filter bad requests before they reach your server. Scanners miss new malware. Firewall rules lag behind new flaws. Layers cover each other, and your patching covers all of them.

Start with the basics if you are on a budget. Close ports you do not use, block direct access to admin areas from unknown addresses and run a malware scan on a schedule. Add the heavier tools as the site grows.

Isolating websites and limiting account permissions

Avoid putting several unrelated sites in one hosting account. If one is compromised, the attacker reaches the rest. Give each site its own account, and on a reseller plan give each client one too.

When you outgrow shared hosting, a VPS or a dedicated server gives you stronger separation from other customers. It also gives you more to manage. You become responsible for the operating system, patching and firewall, so choose it when you can take on that work or pay someone to.

Limit permissions inside your own team too. A person who updates content does not need server root access.

Maintaining encrypted, off site and regularly tested backups

Keep at least one backup copy outside your hosting provider, in a different location and under different credentials. Encrypt it. Keep several versions, because some infections go unnoticed for weeks. If your backup service offers settings that stop deletion for a set period, turn them on.

Then test restores on a schedule. A small site restores in minutes. A large server can take hours, and that time is your real recovery time. Write it down, and compare it with how long you can afford to be offline.

Backups are the one defense that works against almost every incident in this article, including the fire.

How Do You Choose a Secure Web Hosting Provider?

Choose a provider that explains its security practices and backup policy in writing, is open about redundancy, monitors its infrastructure and tells you quickly when something breaks. No provider can promise zero risk. Keep your own backups and account security in place whichever one you choose.

Evaluating server security and infrastructure protection

Ask direct questions. How are customer accounts separated from each other? How fast are security patches applied? Is MFA required for staff? How does the provider handle DDoS traffic and malware found on a customer site?

Look for specifics. The FTC order against GoDaddy turned partly on security claims that did not match practice, so a confident marketing page proves little. We run LiteSpeed and NVMe storage on Intel Dual Xeon servers. Those are performance facts. Whether a provider’s security is good is a different question, and you should ask it separately.

Write the questions down and compare answers across providers. The one that hesitates on a simple question about account isolation is telling you something.

Checking backup policies and disaster recovery capabilities

Ask where backups are stored, how often they run, how long they are kept and whether they sit in a different location from the live servers. The Strasbourg customers learned the hard way that a backup in the same building is not a disaster recovery plan.

Then ask for restore times. Hours or days? If the provider’s own backup is the only copy you have, assume that restore time applies to you and plan around it. Treat the provider’s backup as a second line, and keep your own first.

Ask whether restores are tested. A provider that restores real accounts regularly knows how long it takes, and that knowledge is what you are really buying. Ask for it in writing.

Understanding uptime commitments and redundancy arrangements

An uptime commitment is a contract promise, and its remedy is usually a service credit. It is not insurance against lost business. Read what it covers and what it excludes.

Redundancy matters more than the percentage. Ask about power, network connections and whether your data can live in more than one place. With 25 server locations worldwide, we can place accounts in different regions, and our multi location reseller hosting is built around that idea. A second location is the answer to a facility level failure like the one in Strasbourg.

Ask about planned maintenance too. Some commitments exclude it, which means a scheduled reboot may not count against the promise at all. Know which events the percentage actually measures.

Reviewing security monitoring and incident response practices

Ask whether the provider monitors its network around the clock, how it detects intrusions and what its incident response process is. A provider that answers clearly has thought about it. A provider that answers with a slogan probably has not.

Ask what happens after an incident too. Will you receive a written account, as OVH gave in 2013? Providers that explain failures tend to fix them.

Ask how fast the provider says it can respond to a serious security event, and who makes the decisions. A named on call team is a better answer than a ticket queue.

Ask about their own staff security too. The OVH case started with one employee email account, so a provider’s internal access controls matter as much as its public facing defenses.

Assessing technical support and security incident communication

Security incidents end with a person fixing something, so support quality is a security feature. Test it before you need it. Send a real pre sales question and see how quickly and how specifically you get an answer.

Over more than 20 years and 700,000+ hosted websites, we have learned that customers forgive a problem faster than they forgive silence. Pick a provider that tells you early, even when the news is bad.

Your provider’s security is half of the picture, and your own backups and logins are the other half. To see what our plans cover on the server side, compare our reseller features.

Leave a Reply

Your email address will not be published. Required fields are marked *