Netfilter shows where a public footprint stops proving the business

Summary

• Netfilter Pty Ltd should be judged from the renewal desk, not from the registry page alone. The buyer is paying, if there is still an accountable service behind the records, for managed access, troubleshooting, abuse handling, routing accountability, site continuity and the reduced cost of getting a fault owned.
• The strongest public evidence is bounded. The BTW directory page at https://btw.media/en/directory/netfilter-pty-ltd identifies the company as a published directory entry, APNIC RDAP records identify Netfilter Pty Ltd as registrant for several Australian internet-number resources, and ABN Lookup shows NETFILTER PTY LIMITED as a cancelled Australian Private Company record at https://abr.business.gov.au/ABN/View?abn=53089036019.
• APNIC records describe one resource range as an ISP for independent schools, small to medium enterprise and local government, and one AS number as Somerville managed internet access to Australian schools and corporates. Those descriptions make a service-continuity thesis plausible, but they do not prove current customers, live pricing, present ownership economics or renewal rates.
• The public resource surface includes IPv4, IPv6 and AS records, plus a current abuse-contact trail. It also includes caution signs: ABN cancellation, a cancelled partnership record, older postal addresses, limited public route visibility for one AS, blocked or unavailable current company-site evidence, and no public income statement for the service unit.
• The commercial test is therefore conditional. Netfilter matters if the resource-holder account still supports a live managed-connectivity service whose customers renew because support, restoration, address stewardship and continuity cost less than failure. It matters much less if the public records are only a legacy administrative residue around resources that have moved into another provider's operating stack.

Start With The Renewal

The cleanest way to analyze Netfilter Pty Ltd is to start with a renewal moment. A school business manager, a council information-technology lead or a small enterprise owner looks at an access and support bill. The connection has worked well enough to become ordinary, which is both the provider's achievement and its commercial problem. If the line is quiet, the buyer asks why the organization is not simply using a cheaper national broadband plan, a bundled mobile backup, a cloud-first application setup, an in-house support contractor, or no special service at all. If the line has failed, the buyer asks a harder question: did the provider restore service quickly enough to justify the premium?

That is the economic unit for Netfilter: an accountable resource-holder and operating-continuity account. It is not an IP address, a route table entry, a domain contact or an old company name. It is the service promise that may sit behind them. The customer buys someone to make connectivity usable and dependable at the site level. The customer buys installation coordination, address stewardship, routing accountability, abuse response, supplier escalation, fault diagnosis, and memory of how the site is built. The customer also buys the option to call one accountable party when the classroom management system, municipal office, payment terminal, cloud application, remote desktop, voice service or staff network stops behaving as expected.

That framing matters because the public record is helpful but incomplete. Netfilter's BTW directory page is reachable at https://btw.media/en/directory/netfilter-pty-ltd, and the directory describes an Australian network-resource holder. APNIC's RDAP entity record at https://rdap.apnic.net/entity/ORG-NPL2-AP names Netfilter Pty Ltd, lists it as an organization, ties it to Australian resources, and shows a Somerville Group address and support contact. ABN Lookup's exact active-name search does not show an active current match for the exact company name, while the all-name search at https://abr.business.gov.au/Search/ResultsAll?SearchText=Netfilter%20Pty%20Ltd shows cancelled matches. The detailed ABN record for NETFILTER PTY LIMITED at https://abr.business.gov.au/ABN/View?abn=53089036019 says the ABN was cancelled from 21 April 2023, the entity type was Australian Private Company, and the historical trading name was ISNET - INDEPENDENT SCHOOLS NETWORK. A separate ABN record at https://abr.business.gov.au/ABN/View?abn=14141617410 shows a cancelled partnership record involving Cosys Integration Pty Limited, Netfilter Pty Limited and The Somerville Group Pty Ltd, with The Somerville Group as a trading name.

Those facts are not enough to tell a simple "active ISP" story. The resource records remain active in APNIC, but the Australian business-number records show cancellation. A support email on an APNIC incident-response contact now uses an Atturra domain, but that is not the same as audited proof of an acquisition, live operating unit or customer transfer. Some Somerville and Atturra public website pages are blocked or unavailable through simple retrieval, and no current Netfilter product page, price book, status page or customer case study was found in the source set used for this article. The responsible opening judgement is therefore not that Netfilter is a thriving managed-internet business or a dormant shell. It is that the public evidence proves a resource-holder surface and hints at a managed access history, while the business conclusion depends on operating facts that are not public.

That boundary is exactly why Netfilter is interesting. Many network-company profiles take the existence of internet-number resources as if it were proof of a business. It is not. Resources can be valuable, stale, transferred, hosted by another provider, maintained for legacy customers, or used as administrative anchors inside a wider managed-service arrangement. A public resource-holder record is like a building key on a key ring: it proves that someone once needed access and that an accountability mechanism still exists, but it does not tell us whether the office is full, leased, sublet, vacant or being operated under a new name.

The renewal question forces better discipline. What does the customer actually buy? Why is that unit costly to provide? Which substitutes discipline the price? And what public evidence can or cannot prove about whether the account is still worth funding? Netfilter's records point to independent schools, small and medium enterprises, local government and corporate managed internet. Those are commercial settings where a support-heavy connectivity account can be real. They are also settings where cheaper substitutes are abundant. The article therefore treats public resource evidence as evidence, not as the product. The product, if it exists in live form, is accountable continuity.

What The Public Record Proves

The first layer is corporate identity. ABN Lookup is useful because it is an official Australian public register, but it is also narrow. The record at https://abr.business.gov.au/ABN/View?abn=53089036019 identifies NETFILTER PTY LIMITED, shows the ABN status as cancelled from 21 April 2023, identifies the entity type as Australian Private Company, lists the main business location as NSW 2065, and records a historical trading name, ISNET - INDEPENDENT SCHOOLS NETWORK, from 9 February 2011. That supports a real Australian corporate record and an education-network trading history. It does not support a current ABN-active operating claim.

The all-name ABN Lookup result at https://abr.business.gov.au/Search/ResultsAll?SearchText=Netfilter%20Pty%20Ltd is equally important because it shows three matches, all cancelled. The first is NETFILTER PTY LIMITED. The second is NETFILTER SOFTWARE, a historic business name connected to a different company record. The third is a combined partnership name: Cosys Integration Pty Limited, Netfilter Pty Limited and The Somerville Group Pty Ltd. That third record, detailed at https://abr.business.gov.au/ABN/View?abn=14141617410, says the ABN status was cancelled from 22 October 2024 and lists The Somerville Group as a trading name from 1 July 2002. Publicly, then, Netfilter sits in a cluster of Australian corporate and partnership records around schools-network and Somerville history, but the active-current ABN proof is absent.

The second layer is internet-number accountability. APNIC's entity record for Netfilter Pty Ltd at https://rdap.apnic.net/entity/ORG-NPL2-AP is the strongest source for the resource-holder footprint. It identifies the organization handle ORG-NPL2-AP, names Netfilter Pty Ltd, and lists multiple networks and AS numbers. The associated networks include 115.124.0.0 to 115.124.7.255, 103.105.144.0 to 103.105.147.255, 203.29.67.0 to 203.29.67.255, 124.158.16.0 to 124.158.23.255, and IPv6 ranges 2401:8c00::/32 and 2401:8c01::/32. The associated AS records include AS38307 and AS38858. That is not a trivial public footprint. It is enough to show that Netfilter has been an APNIC resource-holder identity with both address and routing-number records.

The third layer is service description embedded in registry remarks. The APNIC record for 124.158.16.0/21 at https://rdap.apnic.net/ip/124.158.16.0 describes the range as "ISP for Independent Schools / Small to Medium Enterprise (SME) and Local Government." That wording should be used carefully, but it is meaningful because it appears on the authoritative resource record. It supports a historical or administrative link between Netfilter resources and managed internet for schools, SMEs and local government. The AS38858 record at https://rdap.apnic.net/autnum/38858 says "Somerville, Internet Service Provider Australia, Managed Internet Access to Australian Schools and Corporates." That is stronger business-context language than a bare address block would provide. It places the resource-holder surface in the language of managed access, not only technical registration.

The fourth layer is incident and contact accountability. The APNIC incident-response record at https://rdap.apnic.net/entity/IRT-SOMERVILLE-AU lists IRT-SOMERVILLE-AU, a Somerville Group address in St Leonards, and an abuse contact email on the atturra.com domain, with a validation date of 30 April 2026. The Netfilter IP Administration contact at https://rdap.apnic.net/entity/NF22-AU lists a Roseville address and support@somerville.com.au, and the autnum records point routing problems to Somerville support. These details show that the resource records have a live accountability trail in the registry. They do not prove that Netfilter itself is the operating brand, that customers see that name on invoices, or that the support function has the same ownership and staff as it did when the records were first created.

The fifth layer is transfer history. APNIC's transfer log at https://ftp.apnic.net/stats/apnic/transfers/transfers_latest.json records a 23 November 2010 transfer of 115.124.0.0 to 115.124.7.255 from the Association of Independent Schools of Victoria Inc to Netfilter Pty Ltd. That transfer is highly relevant to the education-network thesis. It links a school-associated source organization to Netfilter as recipient for a /21 IPv4 block. It does not prove the customer base after 2010, nor does it show whether those resources still serve independent schools, migrated clients, internal systems or other use cases. It does, however, establish that the schools-network framing is not invented from a name alone.

The sixth layer is public routing visibility. BGP.tools' AS38307 page at https://bgp.tools/as/38307 identifies "ISNet - Independent Schools Network" and reports that this AS is not currently in the global routing table, with zero IPv4 and zero IPv6 prefixes originated in that view. That is not fatal to the business thesis because a resource-holder can have resources not visible under that AS at a given time, and APNIC records can remain valid even when an AS is not actively announcing prefixes publicly. But it is a caution sign. If the only visible proof of live service were AS38307, the case would be weak. The public route view tells us the independent-schools AS record should not be treated as proof of a current routed access base.

These layers prove a bounded proposition: Netfilter Pty Ltd is an Australian corporate/resource-holder identity with historical school-network connections, APNIC-held IPv4 and IPv6 resources, two associated AS records, and Somerville-linked technical accountability. They do not prove present revenue, current customers, margin, service quality, ownership structure, invoice names, customer renewal rates, upstream contracts or active route utilization. A serious business judgement has to stop at that boundary and then ask what operating facts would bridge it.

The Service Unit Behind The Records

If there is a live economic unit behind the Netfilter footprint, it is most likely not "addresses for their own sake." IPv4 scarcity gives address blocks balance-sheet and option value, but a school, council or SME does not renew an annual support relationship because a registry shows a /21. The customer renews because the provider makes the internet account dependable enough that staff can keep working, public services can stay reachable, classrooms can teach, and support pain is lower than it would be under a cheaper substitute.

The 124.158.16.0/21 APNIC wording matters because it identifies the plausible customer categories: independent schools, SMEs and local government. Those segments share a feature that makes managed connectivity valuable. They often have small technology teams relative to the social or commercial importance of their services. A school may have hundreds or thousands of students and staff, classroom devices, cloud learning systems, email, identity services, video, filtering policies, guest access, finance systems and parent communications. A council office may have front counters, development applications, library systems, community services, field staff, telephony, payments and public information. A small enterprise may depend on accounting systems, online sales, customer support, cloud files, security cameras, voice and remote work. None of those buyers wants to be a carrier integration shop.

The paid unit is therefore an accountable service account. It may include a routed connection, managed router, address assignment, firewall configuration, monitoring, filtering, DNS or domain coordination, support desk triage, carrier escalation, incident communication, and restoration work. The exact bundle is not public for Netfilter, so the article cannot claim all those features as facts. But those are the costs and promises that normally make managed internet valuable in the school, council and SME settings named in the resource records.

The accountable service account is costly because it absorbs ambiguity. When a service fails, the problem may be a carrier outage, customer premises equipment, local power, Wi-Fi, DNS, DHCP, a firewall rule, a route announcement, a stale contact, a cloud-provider issue, a content-filter rule, a school bell schedule that makes peak usage predictable, a contractor's cable work, or a building access issue. A cheap broadband plan can hand the customer a modem and a support queue. A managed service has to decide which part of the chain is responsible and who is going to fix it. That decision work is not free. It requires records, trained staff, monitoring tools, supplier relationships and enough technical context to avoid pushing the customer between parties.

The service also carries renewal memory. A provider that has served a school or council for years may know which room has the demarcation point, which old service cannot be interrupted during term time, which contact has authority to approve a change, which support partner owns the firewall, which application breaks when latency rises, and which supplier fails to update tickets unless escalated. That memory is not visible in public records, but it is often the reason a buyer pays more than the cheapest broadband price. If Netfilter's footprint still maps to such a service, the economic value is embedded in the provider's accumulated operating knowledge.

The buyer's willingness to pay is disciplined by substitutes. A school or small enterprise can compare the account against business nbn offers, enterprise fibre, a national incumbent bundle, a specialist managed-service provider, mobile backup, satellite, cloud migration, in-house support or delayed refresh. NBN Co's current Enterprise Ethernet page at https://www.nbnco.com.au/business/enterprise/enterprise-ethernet describes dedicated fibre internet with fast, symmetrical speeds, scalable performance and reliable connectivity for complex needs. Telstra's small-business nbn page at https://www.telstra.com.au/small-business/internet/nbn markets business-grade speeds, 24/7 nbn technical support and mobile network backup on eligible plans. Those public offers do not prove anything about Netfilter's own pricing. They do define the buyer's outside options.

The commercial question is not whether Netfilter has an address block. It is whether a Netfilter-linked account still gives a buyer a better total cost of continuity than those substitutes. The answer would depend on private facts: actual monthly fee, installation charges, speed, included support, backup design, service credits, restoration history, repeat-fault rate, uptime, customer satisfaction, cancellation rate, and how much the customer values not rebuilding the support relationship. The public records identify the battlefield, not the winner.

Why The Unit Is Costly

Managed access for schools, councils and SMEs is expensive because it is a small-scale reliability product. The cost is not only transit or wholesale broadband. The cost is knowing what to do when the ordinary path fails. That starts before the service goes live. A provider has to scope the site, confirm access technology, map existing services, choose equipment, preserve any static addressing needs, schedule installation, work around school terms or council office hours, test failover and document who can approve changes. The customer sees a monthly bill. The provider carries a project.

The first cost category is address and routing stewardship. APNIC records show Netfilter-associated IPv4 and IPv6 resources, including 115.124.0.0/21 at https://rdap.apnic.net/ip/115.124.0.0, 103.105.144.0/22 at https://rdap.apnic.net/ip/103.105.144.0, 203.29.67.0/24 at https://rdap.apnic.net/ip/203.29.67.0, and IPv6 ranges at https://rdap.apnic.net/ip/2401:8c00:: and https://rdap.apnic.net/ip/2401:8c01::. Maintaining resource registration is administrative work. Maintaining correct contact details, abuse response, route policy, customer assignments and reverse DNS is operational work. None of that proves revenue, but it shows why a resource-holder service has costs that commodity broadband customers may not see.

The second cost category is incident response. The IRT-SOMERVILLE-AU record at https://rdap.apnic.net/entity/IRT-SOMERVILLE-AU exists because public internet resources need accountable abuse and incident contacts. For a school or local government customer, abuse handling is not abstract. A compromised host, spam source, open relay, denial-of-service target, student misuse, infected device or misconfigured service can create reputational and operational harm. The customer may not know whether the issue is local policy, a cloud account, a user device or a network problem. A managed provider has to receive reports, triage evidence, contact the right local party, keep the service available where possible and avoid letting one incident contaminate the whole account.

The third cost category is support labor. A public-school or council environment may have non-technical stakeholders who need plain operational answers. Is the issue local or upstream? Should classes continue using a backup link? Does the outage affect payment systems? Is there a security concern? Should staff switch to mobile hotspots? Is the provider waiting on a national carrier? Those answers require staff who understand both technology and customer context. A cheap plan can tell the customer to reboot equipment. A managed account is paid to decide whether rebooting is safe, useful or irrelevant.

The fourth cost category is supplier dependence. A smaller provider or resource-holder account is rarely a complete physical network. It usually depends on upstream carriers, fibre access providers, data centers, domain registrars, equipment vendors, monitoring platforms and sometimes a national broadband wholesale layer. The provider creates value by simplifying those dependencies for the customer. It also assumes coordination risk. If the upstream carrier is slow, if a building access window is missed, if a router vendor firmware issue appears, if a wholesale address record is wrong, or if a carrier fault is intermittent, the managed provider spends labor protecting the customer from fragmentation.

The fifth cost category is redundancy. The renewal case is strongest when the customer can show that service continuity protects real work. That may require a second link, mobile backup, separate provider, spare router, tested failover, alternate DNS configuration, route filtering or separate administrative contacts. Redundancy is expensive because idle capacity still has to be bought, managed and tested. A backup service that is not tested may fail at exactly the wrong time. A backup service that is tested consumes staff time. The provider has to price that discipline somewhere.

The sixth cost category is compliance and governance. Schools and councils do not only buy bandwidth. They care about child-safety policies, privacy, acceptable use, public-service continuity, procurement records, insurance expectations, grant reporting, disaster recovery, data protection and the ability to explain outages. The resource records cannot prove Netfilter's compliance offering. But the customer categories named in the APNIC remarks are categories where documentation and accountability matter. A managed account earns its premium when it reduces the customer's governance burden as well as the outage burden.

The seventh cost category is scarcity and transition. IPv4 resources are finite. Static addressing, legacy allowlists, old cloud firewall rules, VPN peers and external integrations can make a customer reluctant to change providers. That reluctance can help a provider retain accounts, but it can also create responsibility. If the provider is sloppy with address administration or cannot document assignments, the customer inherits migration pain. If the provider is careful, the customer pays partly for continuity of identity on the network.

The public record lets us identify these cost categories, but it does not let us measure Netfilter's cost base. We do not have payroll, carrier invoices, data-center contracts, utilization, support-ticket volumes, gross margin, service-level penalties or customer churn. A buyer considering renewal should ask for those private operating facts in customer-appropriate form: incident history, restoration times, current design diagrams, escalation contacts, backup test dates, route and address assignment records, supplier responsibilities and price comparison against alternatives.

Substitutes Discipline The Price

A Netfilter-linked renewal cannot be judged in isolation because Australian business connectivity has several credible substitutes. The first substitute is a national incumbent plan. Telstra's business nbn page at https://www.telstra.com.au/small-business/internet/nbn says its business internet offer includes business-grade speeds, advanced security, 24/7 nbn technical support and mobile network backup on eligible plans. The page also shows a headline small-business offer around super-fast speeds at $85 per month in the retrieved page text. A small organization that wants simplicity can look at that kind of bundle and ask why a specialist or legacy managed account should cost more.

The answer can be legitimate, but it must be proved. A national incumbent can provide brand confidence, wide coverage, integrated billing, mobile backup and a large support operation. It may not provide the same site-specific memory, school-specific policy knowledge, legacy address continuity or local escalation path that a specialist managed provider can offer. The buyer should not assume either side wins. It should compare the actual time to repair, access technology, backup limits, included support, outage communications, pricing, modem or equipment obligations and the ability to reach someone who understands the site.

The second substitute is NBN Co's enterprise-grade wholesale access through retail providers. NBN Co's Enterprise Ethernet page at https://www.nbnco.com.au/business/enterprise/enterprise-ethernet describes a dedicated fibre service with symmetrical speeds, scalability and reliable connectivity for complex needs. NBN's service-provider list at https://www.nbnco.com.au/business/service-provider-list also reminds buyers that they do not buy directly from NBN Co; they choose a retail or managed provider. For a school, council or SME, the question becomes whether the Netfilter-linked service is an access provider, a managed service layer over such access, a resource-administration function, or a legacy identity within another operating model.

The third substitute is a specialist managed-service provider. The APNIC records themselves point toward Somerville and current Atturra-domain contact evidence. That does not prove current commercial structure, but it does show why the buyer's comparison should include managed IT firms, not only carriers. A managed-service provider may bundle connectivity with help desk, cybersecurity, device management, cloud migration, backup, Microsoft or Google administration, and procurement. If the customer's workload is mostly cloud applications and endpoint support, a broader managed-service bundle may be more valuable than a narrower connectivity account. If the customer's pain is physical access and routed address continuity, the network-specific account may still be superior.

The fourth substitute is cloud relocation. An SME or school can reduce local network dependence by moving applications to cloud-hosted platforms, using identity-as-a-service, outsourcing email, consolidating file storage, using software-as-a-service for finance or learning systems, and using ordinary business broadband at the edge. That does not eliminate connectivity dependence. It changes its shape. The local site still needs reliable access, but the provider may not need to host applications or maintain complex public addressing. A Netfilter-linked account earns its place if it makes cloud access resilient, not if it merely preserves old network complexity.

The fifth substitute is in-house support. A larger school group, council or enterprise can hire staff, buy services directly from carriers, manage its own routers, hold its own addresses, join an exchange, or contract separately for cybersecurity and cloud services. In-house control can lower vendor dependence and improve context. It can also create hidden cost. The organization must recruit, retain, train and cover staff, document the network, manage out-of-hours escalation, maintain vendor relationships and handle incidents. For many customers in the APNIC-described segments, a managed account exists because in-house depth is too expensive relative to scale.

The sixth substitute is mobile or satellite backup. A business can use 4G, 5G or satellite to reduce dependence on a fixed connection. That is valuable, but it is not always equivalent. Mobile backup depends on local coverage, cell congestion, router integration, data caps, antenna placement, failover testing, and whether the applications tolerate latency and address changes. Satellite can be essential for regional locations but may introduce latency, equipment placement and weather sensitivity. The managed provider's value is not merely having a backup; it is proving that the backup works for the customer's actual operating day.

The seventh substitute is delayed purchase or no active dependency. This is the cheapest option and often the most dangerous. A school can postpone a network refresh until the next budget cycle. A council can keep an old service until it breaks. An SME can accept occasional outages as a cost of doing business. Delay disciplines price because it gives the buyer bargaining power. It also creates tail risk. The provider wins the renewal if it can show that the avoided disruption is worth more than the premium.

Because substitutes are credible, the public resource footprint cannot carry the commercial case alone. Netfilter's visible resources and historical schools-network records help explain why a service might matter. They do not show whether the account is priced competitively today. The renewal case needs a simple comparison: what does the current account cost, what does a credible substitute cost, what risks does each option leave with the customer, and what evidence shows that the current account actually reduces those risks?

Resource Evidence Without Overclaiming

The resource evidence is useful when it stays in its lane. APNIC says Netfilter Pty Ltd is the registrant organization for several resources. The organization record at https://rdap.apnic.net/entity/ORG-NPL2-AP lists the networks and AS numbers together. The 115.124.0.0/21 record at https://rdap.apnic.net/ip/115.124.0.0 describes Netfilter Pty Ltd and includes a registration date in 2008 and last-changed date in 2020. The 103.105.144.0/22 record at https://rdap.apnic.net/ip/103.105.144.0 was registered in 2017 and last changed in 2020. The 203.29.67.0/24 record at https://rdap.apnic.net/ip/203.29.67.0 dates to 2008. The IPv6 records at https://rdap.apnic.net/ip/2401:8c00:: and https://rdap.apnic.net/ip/2401:8c01:: date to 2011. That variety suggests the account was not a single accidental entry.

The AS records are also informative. AS38307 at https://rdap.apnic.net/autnum/38307 is named ISNET-AS-AP and described as ISNet - Independent Schools Network. AS38858 at https://rdap.apnic.net/autnum/38858 is named SOMERVILLE-AS-AU-AP and described as Somerville, Internet Service Provider Australia, Managed Internet Access to Australian Schools and Corporates. These are public technical records, not marketing copy. They support a thesis of managed connectivity for education and corporate customers. But because they are technical registry records, they should not be asked to prove customer numbers, revenue or present product features.

The public routing view reinforces caution. BGP.tools reports AS38307 as not currently in the global routing table and with no originated prefixes in that view at https://bgp.tools/as/38307. That single route view does not invalidate APNIC's AS record. It does mean that any claim of current live traffic on AS38307 would require more evidence. The AS may be reserved, inactive, used differently, hidden from that public view, or simply retained as part of an old resource set. A buyer should not confuse AS existence with live service load.

The transfer log adds one important piece of history. The APNIC transfer record at https://ftp.apnic.net/stats/apnic/transfers/transfers_latest.json shows 115.124.0.0/21 moving from the Association of Independent Schools of Victoria Inc to Netfilter Pty Ltd in November 2010. That link is economically meaningful because it connects the resource-holder account to an education-sector source organization, not just a generic internet allocation. It also raises questions. Did Netfilter inherit an education-network service? Did the resources support a managed schools network after transfer? Were schools later migrated to Somerville or another provider? Did the resources become part of a broader managed-service platform? Public records do not answer those questions.

The current contact data raises another important question. IRT-SOMERVILLE-AU shows a 2026 validation for support@atturra.com. That is a live accountability signal inside APNIC. It is not proof that Atturra owns Netfilter, not proof that Netfilter customers are Atturra customers, and not proof that all Netfilter-linked services are live. It does show that the resource records are not completely abandoned. Someone remains tied to the incident-response contact process.

The business lesson is that resource evidence should be read as a map of operating obligations. It tells us there are resources to maintain, contacts to keep current, abuse reports to answer, addresses that may sit in allowlists, and routing records that may need stewardship. It does not tell us whether the associated account earns a margin. It points to the questions an analyst should ask: which resources are currently announced, by whom, for whose traffic, under what customer contracts, and with what support commitments?

For Netfilter, the resource evidence is unusually good at showing a historical managed-access thesis and unusually weak at proving current business performance. That combination should make the article more careful, not less. The public footprint justifies monitoring because it touches education, SME and local-government continuity. It also requires humility because the most decisive facts are private.

Customers, Demand And Retention

The likely customer economics come from the categories named in the APNIC records: independent schools, SMEs, local government and corporates. Each category has a different renewal logic. A school renews if the service reduces learning disruption, protects administrative systems, supports child-safety policies, handles term-time changes and provides help when local technology staff are stretched. A small enterprise renews if the service keeps revenue systems available and reduces owner or staff time spent on technical coordination. A local government office renews if the service supports public access, payment collection, records, staff communications and community-facing systems. A corporate customer renews if the provider reduces the cost of managing multiple sites or legacy address dependencies.

Schools are a particularly plausible historical demand base because the ABN record names ISNET - INDEPENDENT SCHOOLS NETWORK and the transfer log links a block from the Association of Independent Schools of Victoria Inc to Netfilter. The APNIC AS38307 description also names ISNet - Independent Schools Network. But plausible is not the same as current. Public evidence does not show how many schools, if any, still depend on Netfilter-linked resources. It does not show whether the account is billed under Netfilter, Somerville, Atturra or another name. It does not show whether schools migrated away. The article can identify the school-network demand logic, but it cannot assert live school contracts.

The retention mechanism for a school or council service is not only speed. Speed is easy to compare. The retention mechanism is reduced coordination cost. If the provider knows the site, owns the escalation path, maintains address continuity, understands filtering and backup needs, and restores faults without forcing the customer to retell the network history, the customer may renew even when a cheaper plan exists. If the provider cannot show that difference, the customer should reprice the account against NBN-based business plans, national telco bundles and managed IT alternatives.

The same logic applies to SMEs. Many small businesses are now cloud-heavy. They may need less static infrastructure than in the past, but they are more dependent on continuous access. A retail office may need point-of-sale, accounting and voice. A professional-services firm may need remote desktop, document platforms and video. A manufacturer may need cloud ERP and supplier portals. A local services company may need scheduling and customer communication. For these users, a managed access provider is worth paying for when it prevents outages from becoming management crises. It is not worth paying for if it only resells a line without meaningful support.

Local government demand is even more sensitive to public trust. A small council's connectivity may support libraries, community services, development applications, online payments, email, telephony and disaster communications. A managed provider that can explain incidents and support continuity may be valuable. But public procurement also creates price pressure. Councils need defensible value. The provider has to show why its account beats a national carrier, a government panel provider, an NBN Enterprise Ethernet retail offer or a broader managed-service contract.

Customer dependence can create both value and risk. Address continuity, site knowledge and support memory can make the provider sticky. They can also create lock-in. If customers stay only because migration is hard, the account is vulnerable at the next procurement review. If customers stay because the provider repeatedly reduces outages, improves documentation and manages suppliers well, retention is healthier. Public records cannot distinguish between those two forms of dependence.

That is why the renewal pack matters. A buyer should ask for a record of incidents, restoration times, escalation steps, backup tests, address assignments, current support contacts, service boundaries and supplier responsibilities. A provider should be willing to show what it owns, what it resells, what it monitors, how it communicates outages and how it protects continuity. If Netfilter-linked service accounts can show that, the public resource footprint becomes commercially meaningful. If they cannot, the footprint remains an interesting but insufficient proof point.

Regulation, Accountability And Operating Risk

Netfilter's public records create a specific form of accountability risk. APNIC resource records are public because internet-number resources need stable contacts, policy compliance and abuse handling. If an account is neglected, stale contacts can delay incident response, frustrate victims of abuse, complicate routing changes and weaken trust. The fact that IRT-SOMERVILLE-AU shows an abuse contact validated in 2026 is positive evidence of some maintained accountability. The fact that other records still show older Somerville and Roseville details is not necessarily a defect, but it is a reason to ask whether all contact and responsibility paths are current.

The ABN cancellation is a separate risk. A cancelled ABN does not automatically mean the resources are invalid or the service has ended. Corporate groups can restructure, partnerships can end, trade names can migrate, and resource accounts can remain under legacy names while operations move elsewhere. But from a buyer's standpoint, cancellation changes the diligence question. Who is the contracting party? Who invoices the customer? Who owns the service obligation? Who receives abuse reports? Who can approve route changes? Who is accountable if a renewal is signed under a different name from the resource holder?

The Somerville and Atturra signals should be treated as accountability clues. The public contact trail makes it plausible that the resources sit within or near a managed-service environment rather than an unsupported relic. It does not provide enough evidence to write a succession history. The safe conclusion is that any customer or analyst should require a current corporate and support map before treating the resource-holder identity as a live commercial provider.

Operational risk also comes from routing. RFC 7454 at https://www.rfc-editor.org/rfc/rfc7454 describes BGP operations and security practices, including protecting BGP sessions, prefix filtering, max-prefix controls, AS path filtering and related safeguards. RFC 7908 at https://www.rfc-editor.org/rfc/rfc7908 defines and classifies BGP route leaks. These sources are generic technical standards, not Netfilter-specific incident evidence. They matter because any provider maintaining AS numbers, prefixes and upstream relationships has to manage routing risk. A renewal buyer should ask how route filters are maintained, who approves announcements, how changes are tested, and what monitoring exists for reachability.

Operational risk also comes from address scarcity and legacy dependencies. Public IPv4 resources can be embedded in old allowlists, VPN peers, school systems, email reputation, remote-access controls and supplier configurations. A provider that manages those dependencies well reduces migration risk. A provider that manages them poorly creates hidden fragility. IPv6 resources in APNIC records show that Netfilter's account has had IPv6 allocations, but public records do not prove current IPv6 deployment quality. A buyer should ask whether IPv6 is in use, whether dual-stack service is available or needed, and whether customers are being kept on old IPv4-only designs because change is difficult.

Security and governance risk are practical rather than theoretical. Schools, councils and SMEs face phishing, compromised devices, ransomware, data loss, web filtering disputes, device sprawl and supplier risk. The network provider may not own all those controls, but connectivity is often the path through which incidents are noticed. A managed access account earns trust when it can distinguish network issues from endpoint, identity or cloud issues and route the problem to the right owner. It loses trust when every incident becomes a blame exchange.

There is also reputational risk in over-reading public records. If a directory or analyst describes Netfilter as an active service provider without current service proof, readers may infer more than the evidence supports. If the article ignores the APNIC service descriptions and transfer history, it misses the reason the record matters. The correct balance is to say that Netfilter is publicly visible as a resource-holder with historical managed-internet context and maintained contact signals, while present commercial status and operating economics remain unverified.

Unofficial Signals And Their Limits

Unofficial signals are thin for Netfilter. There is no obvious public customer-review pattern, no current public price sheet found, no reachable current Netfilter site in the reviewed material, no public outage history tied to the entity, and no public case study proving live school, council or SME service. That absence should not be stretched into a negative conclusion. Small managed-connectivity providers can be quiet because their customers are institutional, local and long-tenured. They may not market under the resource-holder name. They may be embedded in a larger provider's service desk. They may be mostly legacy support accounts. Quietness is not proof of weakness.

The route-view signal is more concrete but still limited. BGP.tools reporting AS38307 as not currently in the global routing table is a useful caution. It reduces confidence that the independent-schools AS is a visible active public routing platform in that view. It does not say anything definitive about AS38858's current traffic, the use of Netfilter-held prefixes through other origin AS numbers, private connectivity, internal addressing, or customer migration. Public route collectors are snapshots and viewpoints, not financial statements.

The ABN signal is also concrete but bounded. Cancelled ABN records show that the exact historical company and partnership records are not currently active under those ABNs. They do not show the fate of customers, assets, contracts or resources. A cancelled ABN can accompany corporate restructuring, ceased trading, name migration, consolidation into another group or administrative cleanup. The article must not convert cancellation into a claim that all services ended. It can only say that current commercial proof is missing under the exact public ABN records.

The APNIC contact signal is useful but not conclusive. A 2026 validated abuse contact suggests that someone is maintaining an accountability channel. It does not prove a live customer base. Many resource holders keep registry contacts current for governance reasons even if resource use is limited. Conversely, a current contact may sit inside a larger managed-service organization that has live customers under another name. The public record cannot adjudicate that.

The market-signal conclusion is therefore conservative. Netfilter is not a case where public chatter can carry the investment or supplier-dependence thesis. The thesis has to be built from official registry and ABN evidence, then expressed as a set of commercial questions. That is appropriate for a resource-holder profile. The public footprint is real. The business model is plausible. The current economics are not proven.

What Would Change The Judgement

The first fact that would change the judgement is a current service catalogue. If a current provider page, customer-facing PDF or procurement response showed that Netfilter or its successor brand offers managed internet access, school connectivity, local-government connectivity, static addressing, routing support, backup services and defined support levels, the article could move from plausible service economics to product economics. Without it, the article must treat the product as inferred from registry descriptions and historical records.

The second fact is a current contracting map. A buyer needs to know whether the contracting party is Netfilter Pty Ltd, The Somerville Group, Atturra, another related company or a retail service provider using Netfilter-held resources. The ABN cancellation makes this especially important. A current map would not need to be public to be useful for a customer, but a public article cannot assume it. The map should show who invoices, who supports, who owns resources, who can authorize changes and who carries liability.

The third fact is active routing and address-use evidence. Which prefixes are currently announced? Which AS numbers originate them? Which upstreams are used? Which customers or service classes sit behind them? Are the IPv6 allocations used? Are any resources held for migration, disaster recovery or legacy allowlists? A current route and address inventory would turn the public footprint into an operating picture. Without it, the footprint remains a set of resource records.

The fourth fact is renewal evidence. The most important economic proof would be customer renewal and churn data. If schools, councils or SMEs renew because Netfilter-linked support consistently shortens outages and reduces administrative burden, the service has commercial value. If renewals are low, if customers migrate to national providers, if support is slow, or if the only retention lever is migration friction, the value is weaker. Public records do not contain this evidence.

The fifth fact is incident performance. Mean time to acknowledge, mean time to restore, repeat-fault rate, carrier escalation performance, backup success, service-credit history and outage communication quality would show whether the provider earns its premium. For a school or council, one well-managed outage can justify years of support trust; one badly managed outage can end a relationship.

The sixth fact is cost base. A provider can offer high-touch support profitably only if account density, supplier terms, process discipline and staff coverage are strong enough. A small legacy account base can be attractive if customers are stable and support demand is predictable. It can be fragile if a few hard sites consume too much senior engineering time. Public records show resource obligations, not cost structure.

The seventh fact is substitute pricing at the actual site. Telstra's small-business page and NBN Enterprise Ethernet pages provide public benchmarks, but final economics are address-specific and provider-specific. Availability, installation charges, maximum speeds, backup coverage, equipment terms and support scope vary. A renewal buyer should compare the current account against a real quote, not a generic webpage.

The Business Judgement

Netfilter Pty Ltd matters because it shows the exact point where public technical proof stops and business proof must begin. The public proof is not empty. APNIC records identify Netfilter as the registrant for multiple IPv4 and IPv6 resources and two AS records. The records include language pointing to independent schools, SMEs, local government and managed internet access to Australian schools and corporates. ABN Lookup confirms a real Australian company record and a historical school-network trading name, while the APNIC transfer log ties one IPv4 range to a school-sector source organization. Those are serious facts.

But they are not enough to prove a live business. The ABN record for NETFILTER PTY LIMITED is cancelled. The partnership record involving Netfilter and The Somerville Group is cancelled. AS38307 is reported by BGP.tools as not currently visible in the global routing table. No public price sheet, current Netfilter product page, audited segment result, customer count, status page or renewal data was found. The APNIC contact path is maintained, but maintenance of a contact path is not the same as proof of current revenue.

The economic unit should therefore be priced conditionally. If Netfilter-linked resources support a live managed-connectivity service, the customer buys accountability around continuity. That service is costly because it requires resource stewardship, support labor, incident response, supplier escalation, site memory, backup testing and governance. It can be worth paying for when the alternative is a cheaper plan that leaves the customer to coordinate failures alone. It is not worth paying for simply because a registry lists resources.

The substitute set is unforgiving. National providers can bundle business broadband, support and mobile backup. NBN Enterprise Ethernet retail providers can offer dedicated fibre options for complex needs. Managed IT firms can bundle connectivity with cloud, endpoint, identity and security support. Customers can simplify local infrastructure by moving more applications to cloud services. Larger organizations can bring support in-house. The Netfilter-linked account wins only if it reduces total continuity cost better than those options.

That does not make the Netfilter record weak; it makes the valuation test practical. A buyer should not ask whether the public footprint is impressive in isolation. It should ask whether the footprint gives the provider a continuity advantage that a substitute cannot copy cheaply. Historical school-network knowledge might matter if sites still depend on old address assignments, filtering policies, routing assumptions or local support relationships. IPv6 resources might matter if a customer wants modernization without losing continuity. A current incident-response contact might matter if abuse reports and restoration tickets still move through a maintained desk. The same facts matter much less if customers have already migrated to generic broadband, cloud services and unrelated support contracts. Public evidence can identify the possible advantage, but the renewal file has to prove it with current operating detail.

For management, the decision would be to preserve optionality while demanding proof. Do not discard the resource-holder account merely because the ABN record is cancelled; address and route continuity can outlive a particular public business registration. Do not pay a premium merely because the APNIC record is active; technical registration can outlive the business need. The rational posture is to maintain clean contact data, document resource use, verify live customers, test backups, compare substitutes and make each renewal earn its price. That is the difference between an asset-like public footprint and an administratively expensive legacy tail.

The final judgement is therefore a disciplined "watch, verify, then value." Netfilter should remain in the directory because the resource-holder footprint is real and historically tied to meaningful Australian institutional connectivity. It should not be overstated as a proved current ISP business without fresh commercial evidence. The facts that would change the judgement are practical: current contract owner, live service catalogue, active route inventory, customer list or customer category evidence, incident performance, renewal rates, and site-specific substitute quotes. Until those facts appear, Netfilter is best understood as a public accountability surface that may still anchor commercially meaningful continuity work, but does not publicly prove the business by itself.