Summary

  • EcoHosting's public identity joins cleanly. Its website and 2026 service terms name EcoHosting Internet Limitada, RUT 76.764.736-0, at Providencia 1650 office 303; LACNIC registers active AS266855 to the same legal name and address.
  • The network evidence is substantive but bounded. RIPEstat observed AS266855 announcing six IPv4 /24 routes and one IPv6 /32 throughout July 1-15, 2026. Four of those IPv4 routes and the IPv6 allocation are registered directly to EcoHosting, while two IPv4 routes sit within address space registered to Pluschile Internet Limitada.
  • EcoHosting publishes usable product surfaces for shared hosting and VPS, including cPanel, remote power controls, monitoring and backup functions. Its terms also say that unlimited resources remain subject to reasonable-use limits and that backup tooling is not an absolute guarantee of complete or successful recovery.
  • The commercial site advertises ticket support around the clock and a 99.9% availability reference, but the master terms do not turn those headlines into response, resolution or compensation promises by themselves. Buyers need a service schedule that defines measurement, recovery, locality, escalation, evidence and exit.

The legal company and the network record point to the same place

A regional hosting provider should be easy to identify before a customer entrusts it with a domain, mailbox or application. EcoHosting clears that first hurdle with unusually direct public evidence. Its company page names EcoHosting Internet Limitada, publishes RUT 76.764.736-0, gives a commercial address at Providencia 1650 office 303 in Santiago and lists a central telephone number. The current service terms repeat the legal name, tax number and address.

The LACNIC record for AS266855 independently names ECOHOSTING INTERNET LIMITADA as registrant. It marks the autonomous system active, records its registration on July 17, 2018 and gives the same Providencia address. The telephone attached to the registrant is the number published on the commercial site. That is a useful identity join: the party selling the service and the party recorded as responsible for the network are not merely connected by a similar brand.

The BTW directory profile supplies the commissioning anchor. It places the company in Chile, traces it to a LACNIC member record and associates it with managed network, cloud, data-centre, colocation and hosting activity. Those labels frame the questions; they do not certify each service or establish which physical assets EcoHosting owns.

The distinction matters because the company website makes broad claims about operating since 2011, serving more than 10,000 customers and managing more than 30,000 websites and mailboxes. Those figures are provider-authored and were not corroborated by a customer list, audited report or public service dataset. A buyer can rely more heavily on the legal and registry join than on the scale headline. The resulting conclusion is narrow but valuable: EcoHosting is a traceable Chilean counterparty with a traceable network identity.

AS266855 is real network evidence, not an uptime certificate

EcoHosting does not ask customers to take the phrase "local infrastructure" entirely on faith. LACNIC registers the active IPv4 block 45.239.108.0/22 to the company. A /22 contains 1,024 IPv4 addresses. LACNIC also registers the active IPv6 allocation 2803:12a0::/32 to EcoHosting. Both records date from July 2018 and name the same administrative, technical and abuse contact as the autonomous-system record.

Routing observations show that these resources are not dormant labels. RIPEstat's announced-prefix record for AS266855 observed seven routes continuously across its returned July 1-15, 2026 window: four /24 routes covering EcoHosting's 45.239.108.0/22, two /24 routes at 192.140.58.0 and 192.140.59.0, and the 2803:12a0::/32 IPv6 route. Each IPv4 /24 contains 256 addresses.

The route list also reveals an operating boundary that a customer should not flatten. LACNIC registers the larger 192.140.56.0/22 allocation containing those two visible /24s to Pluschile Internet Limitada, not EcoHosting. The two companies share the same named network contact in the records, but the public documents reviewed here do not define their corporate or service relationship. EcoHosting may have a legitimate delegated use or operating arrangement; the evidence does not explain who controls allocation records, route changes, abuse handling or customer support for that portion of the address space.

There is one positive route-security signal. RIPEstat's RPKI validation returned valid for AS266855 and the EcoHosting /22, with a maximum route length of /24. This lets networks using route-origin validation check that AS266855 is authorised to originate the four more-specific routes. The same check returned unknown for the IPv6 /32, meaning the capture found no validating route-origin authorisation for that prefix.

These records establish number-resource responsibility, visible routing and one authorisation control. They say nothing about packet loss, latency, transit diversity, DDoS capacity, route-change governance, server occupancy or application availability. A provider can announce every expected prefix while a disk array, hypervisor, DNS service or support queue is failing. The right procurement response is to ask for a customer-specific path: assigned addresses, origin ASN, upstreams, facility, switching and firewall layers, monitoring point, failover design and the team authorised to change each layer.

The product surface replaces tasks, but it also divides responsibility

EcoHosting's current hosting catalogue is concrete enough to show what a small-business customer is buying. Plans combine website storage, corporate mailboxes, SSL, cPanel, MySQL, application installation and a site builder. The page names LiteSpeed Enterprise, BitNinja and JetBackup, and it distinguishes entry, small-business and higher-capacity plans by storage, mailboxes, sites and compute resources.

That bundle removes several pieces of manual work. A small company can register a domain, publish WordPress, create mailboxes, manage a database, obtain a certificate and request restoration without assembling separate suppliers. The automation is economically meaningful precisely because the customer does not need to operate every underlying component.

The VPS offer exposes a different control boundary. EcoHosting describes immediate provisioning, remote restart, shutdown and power-on through a web panel, SSD storage, optional cPanel, two full-server images per week, external backups and node monitoring through PRTG with email and SMS alerts. These are actions and mechanisms a buyer can test. They are more informative than a generic cloud label.

The control plane still leaves important questions unanswered. The public pages do not define role separation, multifactor authentication, API access, audit-log export, emergency-access controls, snapshot consistency, backup encryption, restore time or how a customer's action is reconciled with an operator action. Nor do they show whether the hosting and VPS layers share DNS, storage, authentication, network or support dependencies.

EcoHosting's own terms supply the necessary caution. "Unlimited" storage, sites, mailboxes, databases or transfer remain subject to normal and reasonable use. The provider may apply limits to CPU, memory, processes, storage, inodes, transfer, mail, connections, snapshots, restorations, ports and other parameters. A buyer should therefore treat the plan name as a commercial tier, not as a capacity model. Expected traffic, mailbox volume, database work, concurrent processes, storage growth and burst behaviour should be written into the accepted order with a warning and migration path before limits become service interruptions.

A 99.9% headline becomes useful only after its denominator is known

EcoHosting's commercial pages refer to 99.9% uptime and ticket support available 24/7. Both can be meaningful. Neither is self-defining.

The master terms say that an SLA, credit, compensation, commercial guarantee or availability commitment exists only when a current policy expressly applies it to the contracted service. They also say that a 24/7 support reference means general availability through the channels EcoHosting defines for that service; on its own, it does not establish a maximum response or resolution time. In the absence of a more specific policy, the terms do not promise an exact response or resolution time.

For a 30-day measurement period, 99.9% availability would permit about 43 minutes of counted unavailability. But that arithmetic is useful only after the contract identifies the measured component and clock. Is the commitment for network reachability, a physical server, the hosting control panel, DNS, mail, the customer's web application or the whole service? Is measurement taken from EcoHosting's monitor or from outside AS266855? Are maintenance, attacks, software faults, customer configuration and third-party failures excluded? Does the clock stop at workaround, partial restoration or complete recovery?

A service schedule should answer those questions and specify evidence. Monthly uptime reports, incident timestamps, alert history, route observations and ticket records should reconcile. If compensation is offered, the schedule should state the claim window, calculation and remedy. Without those definitions, 99.9% remains a useful signal of intent but an unreliable basis for valuing downtime risk.

Chilean operation does not locate every copy of customer data

EcoHosting consistently presents itself as a Chilean operator. The legal entity, office, ASN, IPv4 allocation and IPv6 allocation all support that description. The hosting pages also say the infrastructure and support are in Chile. That is relevant to contracting, local contact and network latency.

It does not establish the location of every data class. Shared-hosting files, virtual-machine disks, snapshots, external backups, mail queues, DNS data, monitoring records, support attachments, account credentials and billing records may follow different paths. The public pages reviewed for this article do not name the production facility, backup site, subprocessor list, inter-site distance or storage operator for each product.

The master terms are explicit about the remaining customer responsibility. JetBackup, snapshots, restorations and similar mechanisms are described as an available tool or operational layer rather than an absolute guarantee of retention, integrity or successful recovery. EcoHosting says it does not guarantee in every case that copies remain available, cover all data or restore completely, and it requires customers to keep their own backups when continuity demands it. Expiration, non-payment or cancellation may also lead to lost access, content deletion or an inability to recover data.

This makes restoration and exit part of the buying test, not paperwork for the end of the relationship. The customer should map each data class to a primary location, backup location, operator, retention period, encryption key owner and deletion process. It should restore a representative site, mailbox, database and virtual-machine image into an isolated environment, then record recovery point, recovery time and missing dependencies. It should also export data and configuration in usable formats before testing cancellation or migration.

"Data in Chile" is valuable only when the contract identifies which data, in which Chilean location, under whose control and with what recovery evidence.

Local support has to connect the ticket to someone who can act

For a small business, EcoHosting's local support proposition may matter as much as its ASN. The site publishes a Chilean telephone number and sales address, while the client portal exposes tickets, a knowledge base and a network-status route. The commercial catalogue advertises continuous ticket support. Those are visible entry points, and the terms commit EcoHosting to maintaining official channels appropriate to the service.

The gap is authority. A ticket handler may be able to reset an account but not change a route, inspect a hypervisor, restore a backup or enter a facility. The public material does not publish severity definitions, on-call roles, escalation contacts, acknowledgement targets, update cadence or restoration distributions. It also does not explain when Pluschile, a software vendor, a facility operator or another supplier must act.

Buyers should test the support chain before placing a critical workload. Submit a low-risk incident through the prescribed channel, verify the timestamp and severity, ask for escalation and observe who can diagnose each layer. A tabletop exercise should cover a blocked address, failed website, lost mailbox, corrupted database, unreachable VPS and suspected route incident. The response record should identify the owner, evidence used, decision authority, handoff and next update time.

That exercise measures the labour behind the interface. Locality can shorten the commercial conversation, but it does not automatically shorten restoration. What matters is whether the person receiving the ticket can reach the person controlling the failed component, especially outside ordinary office hours.

Procurement should begin with a measured service map

EcoHosting has already supplied more public proof than a name alone. The legal identity is consistent, AS266855 is active, directly registered IPv4 and IPv6 resources are visible, expected prefixes are routed, one key IPv4 origin is RPKI-valid, and the product pages expose real hosting and VPS controls. Those facts justify serious evaluation.

They also show how to run it. First, bind the order to EcoHosting Internet Limitada, the exact plan, assigned resources, facility, origin ASN, any Pluschile dependency and every third party that can interrupt delivery. Second, test provisioning, role access, remote power actions, monitoring, backup and restoration with timestamps. Third, fail a permitted dependency and measure detection, escalation, workaround, recovery and communication. Finally, export the workload, mail, DNS, logs and account data, stop billing and verify deletion.

The acceptance record should become part of the service schedule. It should define capacity thresholds, availability measurement, recovery objectives, data locations, support clocks, evidence rights, remedies and exit assistance. That is the point at which network presence becomes operating assurance.

EcoHosting can credibly say that there is a Chilean company and network behind the storefront. The public record does not prove the outcome of the next disk failure, route change, restore request or overnight ticket. A buyer should credit the traceable footprint, then make the service earn the rest through a measured workload and a contract that names who acts when the automation stops.