FOP "Makurin Stanislav Volodimirovich" makes a local connectivity bill depend on upstream routes and field repair

Summary

• The network is operating. AS42510 had all 17 of its overlapping IPv4 announcements visible to every one of 327 IPv4 peers in a RIPE RIS snapshot on 10 July 2026. Those announcements represent 3,584 unique IPv4 addresses, not 17 separate address blocks, and no IPv6 route was visible.
• JetNet presents a current regional fibre service across named locations in Kyiv and Kyiv Oblast. Its site accepts connection requests, describes GPON launches in villages including Kiyliv, Hvozdiv and the Bezradychi area, sells customer power equipment and says XGS-PON deployment is in progress. This supports current operation, but not a complete plant map or proof that every advertised location is reached by owned fibre.
• The exact operator boundary requires a downgrade. RIPE registers AS42510 to FOP "Makurin Stanislav Volodimirovich", while JetNet calls itself a group of companies, tells customers to send bank transfers to Multimedia Network LLC, and UA-IX lists that LLC as the AS42510 member. Public records also connect Stanislav Makurin to two registered telecom companies. The FOP is clearly a network-resource identity, but its present responsibility for billing, plant, staff and service contracts cannot be isolated from the group.
• Logical route diversity is better documented than physical resilience. Current records show two transit-facing networks and three exchange ports, including declared 40 Gbps ports at DTEL-IX and Giganet and a 10 Gbps port at UA-IX. Yet there is no public fibre-route plan, facility list, power-autonomy duration, tested failover result, utilisation series, crew roster, spare-parts policy or restoration record. The evidence supports a regional ISP network with a strong routed presence and a weakly documented physical recovery design.

The name on the route is not the whole operator

The cleanest fact about this company is also the easiest to overstate. The RIPE registration for AS42510 names FOP "Makurin Stanislav Volodimirovich" as the registrant, identifies the autonomous system as WDM-AS and records the resource as active. The autonomous system was registered in March 2007 and its record was changed in October 2025. Its organisation address is in Kyiv. That is strong evidence that the sole proprietorship is the registered holder of a long-lived internet routing identity.

It is not, by itself, proof that the FOP is the only company installing customer drops, sending invoices, employing technicians or owning the optical line terminals behind the JetNet name. An autonomous system registration answers who holds the routing resource. It does not provide an asset register or reproduce the customer contract.

JetNet's own pages point to a wider operating arrangement. The company description calls JetNet a group of companies active in the Ukrainian telecom market since 2005. It says the group invests in its own network in Kyiv and Kyiv Oblast and serves households, businesses, educational institutions and other customers. Most decisively, the payment page tells customers making a bank transfer that the recipient is Multimedia Network LLC, company code 38730741. The UA-IX participant list also places AS42510 and the JetNet website under Multimedia Network LLC rather than under the FOP.

Public company data reinforce the distinction. An Opendatabot profile for Stanislav Makurin connects him to Multimedia Network LLC and Jet Net Telecom LLC. It lists the latter's principal activity as wired telecommunications and the former's as other telecommunications. The same page describes the current FOP activity under real-estate leasing and operation, not telecommunications, while warning that name-only records can produce mistaken identity. A YouControl profile of Multimedia Network reports it as registered, with telecom and telecom-construction activities. A separate Jet Net Telecom profile says that company was not in termination as of 9 July 2026 and had wired telecommunications as its main activity.

These records do not sever the FOP from the network. AS42510 remains registered to it, and company records connect Makurin to the firms visible around JetNet. They do change the correct wording. The public record supports "the JetNet network associated with AS42510 and the Makurin FOP" more strongly than it supports "the FOP alone operates every layer of JetNet." The distinction matters during a fault. A customer needs to know which party owns the damaged drop, which controls the upstream contract, which holds the batteries and which dispatches the technician. The available pages do not allocate those responsibilities among the FOP and the companies.

That is the first downgrade. The exact entity can be assessed as a current network-resource holder within an active regional ISP operation. It cannot be presented as a self-contained operator whose ownership boundary is fully visible.

A regional footprint is visible, but the plant map is not

JetNet's public sales surface is more specific than a generic claim to cover Kyiv. Its location sitemap contains Ukrainian-language service pages for a cluster of places in and around the capital. They include Osokorky and residential developments in Kyiv, and villages or settlements across the wider Kyiv area such as Hnidyn, Shchaslyve, Voronkiv, Petropavlivske, Vyshenky, Protsiv, Kiyliv, Hvozdiv, Novi Bezradychi, Stari Bezradychi, Mali Dmytrovychi, Velyki Dmytrovychi, Romankiv, Pidhirtsi, Khodosivka, Khotivka and Novosilky. The sitemap was still being updated in June 2026.

Individual pages make the footprint concrete. The Kiyliv service page accepts a connection order for the village in Boryspil district and describes xPON access. The Osokorky page does the same for the private-house sector of Kyiv. The Khotiv page names a village southwest of the city. The spread is consistent with a regional operator whose access network reaches both suburban housing and villages, rather than with a single-building reseller.

The operator also published two useful launch claims. One coverage announcement says GPON service began in Hvozdiv, Mali Dmytrovychi, Velyki Dmytrovychi, Stari Bezradychi and Novi Bezradychi. Another Kiyliv announcement says unlimited GPON service launched there. Both are dated 24 July 2025 on the current site. Together with the live order forms, active customer portal and current contact details, they are stronger operating evidence than an old business listing.

But a list of places is not a route map. It does not show where the feeder fibres run, where a handhole or splitter sits, whether the operator leases strands from another owner, or where traffic crosses from the access network into metro transport. A service page can represent dense coverage, a single street, a housing development, a partner arrangement or an address-by-address build. The site's coverage map page invites a visitor to select an area but publishes no readable topology, fibre length, node addresses or ownership boundaries in its text.

The difference is material because geography creates shared risks. Several named settlements lie south or southeast of Kyiv, while others are north of the city. A broad footprint could be supported by several independent aggregation corridors. It could also converge onto one metro site or one cable crossing. A route from Kiyliv or Protsiv may share a bridge, roadside duct or pole line with the route from nearby Vyshenky. A backhoe, damaged pole, flooded chamber or attack on one facility could then affect more places than the sales map suggests.

The public claim of investment in an "own network" also requires care. Ownership may apply to customer drops, passive splitters and selected feeders while long-haul or metro segments are leased. It may apply to active equipment while ducts and poles belong to municipalities, utilities or another carrier. No published asset inventory separates owned fibre from indefeasible-use rights, leased capacity or partner access. No public record identifies a ring around Kyiv or a second route into each village.

The regional-ISP classification therefore survives, but with a defined boundary. The named locations, continuing site activity and routed customer population support a regional service footprint in Kyiv and Kyiv Oblast. They do not establish ubiquitous coverage, complete control of the physical path or regional resilience.

Seventeen announcements describe 3,584 addresses, not seventeen networks

The strongest independent operating signal comes from the routing system. In the RIPE routing-status view for AS42510, all 327 IPv4 RIS peers in the 10 July 2026 snapshot saw at least one route from the autonomous system. The latest route observation was current to the snapshot hour. No IPv6 prefix was visible to any of 321 IPv6 peers.

The announced-prefixes view listed 17 IPv4 routes over the preceding two weeks. That number can mislead because many announcements overlap. The network originates 176.103.64.0/21, its two /22 components and four /23 components. It similarly originates 193.106.56.0/22 together with two /23s and four /24s, plus 193.200.68.0/23 and its two /24s. Counting every aggregate and more-specific route produces 17 announcements; counting unique address space produces 3,584 IPv4 addresses.

An aggregate is a compact direction sign. A more-specific route can steer part of the traffic differently. AS42510's pattern is therefore evidence of routing control, not of 17 physically separate customer networks. The same fibre port can carry all of them. Conversely, several circuits can carry one aggregate. Address and route counts cannot reveal the number of optical line terminals, access splitters, villages, subscribers or gigabits of usable capacity.

Independent summaries agree on the basic scale. Hurricane Electric's AS42510 profile lists 17 originated IPv4 prefixes, no IPv6 prefix and 3,584 originated IPv4 addresses. IPinfo's AS42510 page also reports 3,584 IPv4 addresses and classifies the network as an ISP. Cloudflare Radar identifies WDM-AS or JetNet and has recently displayed an estimated customer population of about 15,000 users. That estimate is useful for order of magnitude, not as a subscriber count. One household line can serve several people, and address sharing can place many users behind one public address.

Route-origin security is a genuine positive. RIPE's validator marks the three base announcements 176.103.64.0/21, 193.106.56.0/22 and 193.200.68.0/23 valid for origin AS42510. Hurricane Electric reports all 17 visible routes as RPKI-valid. That helps other networks reject an unauthorised origin when they perform route-origin validation.

RPKI answers a narrow question: is this autonomous system authorised to originate this prefix? APNIC's RPKI explanation makes clear that route-origin authorisation is about verifiable number-resource holdings and routing intentions. It does not authenticate the full path, keep an edge router powered, detect a damaged fibre or guarantee that a backup route has enough capacity. AS42510 looks orderly at the route-origin layer. That should not be converted into a broader reliability claim.

IPv6 is the clearest missing capability. No current IPv6 announcement appears in RIPE, Hurricane Electric or IPinfo. The website discusses modern optical access and a transition toward XGS-PON, but access speed and internet-protocol support are separate layers. A 10-gigabit optical terminal can still deliver IPv4-only service. Without a visible IPv6 allocation and route, current native IPv6 service cannot be credited.

Logical diversity is real; physical diversity is not demonstrated

AS42510 is not a simple one-upstream edge in the public routing view. Hurricane Electric and IPinfo both identify Eurotele-Plus, AS6768, and UnderNet, AS41435, as upstreams. A RIPE BGP-state snapshot shows AS6768 and AS41435 immediately before AS42510 on many paths to aggregate announcements, after consecutive prepending is treated as one origin. More-specific announcements are also visible through exchange and peering paths. This is materially better evidence than a single observed provider.

The exchange footprint is also current and substantial for a regional ISP. PeeringDB's network record classifies the network as regional cable, DSL or ISP access, reports an open peering policy and places its declared traffic band at 20 to 50 Gbps. Its exchange attachment records show three operational IPv4 ports: 40 Gbps at DTEL-IX, 10 Gbps at UA-IX and 40 Gbps at Giganet. Each record says the network uses the exchange route server. The entries were updated between April and July 2026.

These facts support three limited conclusions. First, the network has more than one logical way to receive traffic. Second, it can reach local or regional networks without sending every packet through paid international transit. Third, the network team has kept public interconnection records current. The large list of adjacent autonomous systems in the RIPE neighbour view is consistent with route-server peering at multiple exchanges.

They do not prove three physical circuits on three independent routes. Two exchange ports can terminate on the same router, in the same rack, powered from the same distribution board. They can be reached over wavelengths in one cable. A transit and an exchange can enter a building through the same duct. A 40 Gbps attachment can be a logical service delivered over shared transport rather than a separate fibre pair. PeeringDB publishes interconnection declarations, not construction drawings.

There is also an ownership clue in the exchange data. PeeringDB calls the network FOP Makurin and uses JetNet LLC as an alternate name, while UA-IX names Multimedia Network LLC for AS42510. Those labels can coexist operationally, but they underline why the exact contract and asset boundaries matter. A route may be registered to one party, an exchange cross-connect contracted by another and a retail bill issued by a third company in the same group.

The operator's strongest reliability statement appears on the JetNet home page. It says the network uses backup channels and that all communication channels switch automatically during faults without loss of connection or service quality. Current routing evidence makes automatic rerouting plausible at the internet edge. It does not substantiate the absolute "without loss" element. BGP failover can take time. A backup path may have different latency or less capacity. A local feeder cut before the edge routers would not be repaired by an upstream route change.

To establish physical diversity, an operator would need to disclose more than carrier names. The useful evidence would show that at least two circuits enter through separate ducts or pole routes, terminate on separate edge equipment, have independent power, and are exercised under load. It would identify which prefixes and services fail over, how long convergence takes, and whether the backup carries peak demand without congestion. None of that is public.

The fair grade is consequently split. AS42510 has good logical diversity for its apparent size: two transit-facing paths, three exchanges and numerous visible peers. Physical route diversity remains unverified. A customer can take some comfort from the first fact but should not infer the second.

xPON moves the power problem; it does not remove it

JetNet's access claim is based on xPON, a family of passive optical network technologies. Its home page says every tariff uses optical access and can continue during a mains outage. The FAQ gives the more accurate version: JetNet says it provides backup power at nodes containing active equipment, while the customer must power the optical terminal and router with a mini-UPS, battery station, generator or inverter.

This architecture has a real advantage over an access network filled with powered street switches. The ITU description of passive optical distribution explains that a simple optical distribution network is a single-rooted point-to-multipoint tree of fibre and passive components. Downstream signals travel from an optical line terminal to splitters and then to optical network units at customers. Passive splitters do not need grid power in the field. That reduces the number of active sites that must be backed up.

It does not make the service independent of electricity. The optical line terminal at the root needs power. So do aggregation routers, exchange switches, monitoring systems and upstream facilities. At the customer end, the optical network unit and Wi-Fi router need power. If any required active node exhausts its battery, the optical path can remain physically intact while service stops.

JetNet's own equipment catalogue makes the customer-side dependency tangible. It lists a 59.2 Wh mini-UPS intended for routers and optical terminals. The page describes a zero-millisecond change between mains and battery for the product. That is evidence that JetNet anticipates customer demand for backup power. It is not a measured runtime for every combination of optical terminal, router, battery age and Wi-Fi load.

National guidance supplies context, not operator-specific proof. Ukraine's Ministry of Digital Transformation says in its November 2025 xPON guidance that xPON can support service for more than 72 hours without grid electricity when the provider's active nodes and the customer's equipment are properly powered. It reports that 1,200 providers offered the technology and that xPON's share of fixed subscribers had risen from 30 percent at the end of 2021 to 52.5 percent in 2025. Kyiv Oblast was among the regions where more than 70 percent of networks were described as prepared for blackouts.

The key word is "can." JetNet does not state the tested autonomy of each node. It does not give battery watt-hours, normal load, replacement dates, generator coverage, fuel stocks or the number of hours maintained during the latest outage. It does not say whether the two transit paths and three exchange services depend on the same powered facility. A promise that the network does not depend on electricity is therefore too broad. The narrower FAQ claim, backed nodes plus customer-supplied power, describes the engineering reality more faithfully.

Power resilience is also a maintenance problem. Batteries lose capacity with age, temperature and cycling. A long outage requires charging or replacement, and a generator requires fuel, safe placement and someone able to reach it. Repeated cuts can begin before batteries have recharged fully. An operator serving dispersed villages has to monitor the state of multiple sites and decide where scarce mobile power or technicians go first.

The correct assessment is positive but conditional. Passive optical access reduces intermediate power exposure, JetNet acknowledges backup at active nodes, and the catalogue supports customer preparation. The duration and breadth of that resilience are not publicly measured.

XGS-PON is a capacity plan, not yet a capacity result

The JetNet FAQ says the company is introducing XGS-PON and that it will enable symmetric service at up to 10 Gbps. The tense matters. "In the process of commissioning" is an expansion statement, not confirmation that every location, optical line terminal, splitter path or retail account can use 10 Gbps today.

There are at least four capacity layers between that headline and a customer's sustained throughput. The first is the optical access port shared among customers on a PON tree. The second is aggregation from optical line terminals toward the metro network. The third is the interconnection and transit edge. The fourth is the capacity of the remote service the customer is using. A 10 Gbps-capable optical standard at the first layer does not guarantee 10 Gbps through all four.

The public interconnection records are useful but also easy to add incorrectly. PeeringDB declares two 40 Gbps exchange ports and one 10 Gbps port. Those nominal port speeds cannot simply be summed into 90 Gbps of internet capacity. Exchange ports may carry overlapping peer sets and local traffic, not full transit. Their use varies by time of day. A provider can have a 40 Gbps physical port while purchasing or engineering for a smaller sustained load. PeeringDB's 20 to 50 Gbps traffic band is self-reported and broad.

Likewise, 3,584 public IPv4 addresses do not set the subscriber ceiling or the access speed. Carrier-grade address translation can place many households behind fewer addresses. Static addresses can be assigned to businesses or infrastructure. Some space can be unused. The 15,000-user estimate shown by Cloudflare is inferred from measurements, not an account ledger.

JetNet's public residential page currently requires a visitor to select a settlement before tariff details are populated, and the text accessible without that interaction does not expose a stable price-and-speed table. The business page asks for an address and a commercial request. It would be unsafe to assign a current tariff, contention ratio or service-level commitment from those pages. Promotions offering free months indicate active customer acquisition, but not margin or capacity headroom.

Installed capacity and usable capacity should therefore be kept separate. The active AS, exchange ports, GPON locations and equipment catalogue show meaningful installed infrastructure. The company has also signalled an XGS-PON upgrade. What remains unknown is busy-hour utilisation, oversubscription, optical split ratios, the number of upgraded ports, transit commits, cache capacity and how much spare capacity remains during failover. Without those figures, the network can be described as expanding, not as proven non-congested at its advertised maximums.

A passive tree can still have one vulnerable trunk

PON lowers field-power requirements because splitters are passive, but the usual topology also concentrates users on shared components. A cut near one household may affect one drop. A cut in a distribution segment may affect a street. A cut in the feeder before a high-level splitter can disconnect many branches. A failure at the optical line terminal can take down every customer attached to its ports.

This is why "fibre" and "resilient" are not synonyms. The ITU summary on PON protection says redundancy is generally not fundamental to point-to-multipoint PONs, although high-density residential and business services can justify added protection switching. The GPON general characteristics describe dual-parenting designs with separate optical line terminals, potentially in physically diverse locations, to protect against equipment, power-supply and site failure.

No public JetNet page claims such a protected access design. There is no evidence of Type B or Type C PON protection, dual optical line terminals for the same customer group, redundant high-level splitters or two feeder paths into a village. The claim of reserved channels appears to concern broader connectivity and may not cover the access tree.

That gap defines the main physical failure path. Consider a household in Kiyliv. Its service needs a powered optical terminal and router in the home, an intact drop, an intact distribution route through one or more splitters, an intact feeder, a powered optical line terminal, powered aggregation, a path to the internet edge and at least one working upstream or exchange. Logical diversity at AS42510 protects only the later part of that chain. It does not bypass a cut customer drop or a dead optical line terminal.

The location pages also imply different construction environments. Private-house districts can involve aerial spans, roadside routes and longer drops. Apartment or residential-complex service can depend on building access, risers and common telecom rooms. Village feeders can run longer distances between nodes. Each setting changes the likely fault and the permission required to repair it. None of the pages says whether poles, ducts, chambers or building routes are owned by JetNet or another party.

A ring topology could protect aggregation between settlements, but no ring is published. Even a ring is only as diverse as its route. Two fibres placed in the same duct can fail together. Two directions that meet at one powered cabinet can share a node failure. The useful evidence would identify independent paths and common points, not merely draw a loop.

Until such evidence exists, the access network should be treated as operational fibre with unknown feeder protection. This is not a claim that it lacks redundancy. It is a refusal to infer protected plant from the word xPON or from diverse internet-edge routes.

Field repair is the capacity that route collectors cannot see

Every fibre network eventually depends on people who can find and repair a fault. JetNet advertises a quick technician visit and says its support responds quickly. The home-service metadata promotes 24-hour support and installation within 24 hours, while the contact page provides several telephone numbers and an email address. These are current customer-facing signals, but they are not restoration statistics.

A credible field operation needs more than a number to call. It needs monitoring able to distinguish a customer power problem from an optical loss, line-card failure, feeder cut, upstream outage or congestion event. It needs accurate route records, access to buildings and roadside plant, optical test equipment, splice machines, compatible modules, spare optical terminals, cable and closures. It needs a crew close enough to reach the site and authorised to work there.

The regional footprint makes dispatch harder. A technician based in Kyiv may face travel time to a village, road restrictions, weather, fuel constraints or security conditions. Several simultaneous faults can exhaust a small team. A cut on leased infrastructure may require escalation to another owner. The customer-facing company may diagnose the failure but have no legal right to open the chamber or splice the cable.

The group boundary returns here. JetNet says its technicians visit customers, but it does not say which legal entity employs them. Multimedia Network receives bank transfers and appears at UA-IX. AS42510 remains registered to the FOP. Jet Net Telecom has wired telecom as its principal activity. Without a service contract or published responsibility matrix, a reader cannot allocate repair authority among them.

No public page gives crew count, operating bases, shift coverage, average response, mean time to restore, first-time fix rate or spares inventory. The 24-hour support claim may describe telephone availability rather than round-the-clock field dispatch. "Quick" has no defined threshold. Nor is there a public priority policy explaining whether a failed village feeder, business circuit and single household are handled differently.

Ukraine's wider telecom context raises the value of those missing figures. The 2026 Rapid Damage and Needs Assessment estimates $2.5 billion of damage and $2.7 billion of losses across telecommunications, digital and media through the end of 2025. It attributes losses partly to increased repair and backup-generation costs, and names rapid repair and expansion of backup power and satellite connectivity among immediate priorities. Kyivska Oblast is one of the areas where sector losses are concentrated.

The national figures cannot be assigned to JetNet and do not prove that its plant has been damaged. They explain why a regional provider's repair capacity is not a minor service feature. It is part of the network's usable capacity. An unlit spare fibre without a crew, access permission or compatible optics may not shorten an outage. A well-stocked team with reliable route records can sometimes restore service faster than a more elaborate but poorly maintained design.

The current grade for local support labour is therefore plausible but unmeasured. The operator solicits orders, publishes support contacts, sells equipment and advertises technician visits. Its field depth and recovery performance remain private.

The retail bill purchases a chain of dependencies

A monthly broadband payment does not buy one asset. It contributes to the customer drop, optical terminal, splitters, feeder, optical line terminal, aggregation, exchange ports, transit, addresses, monitoring, support, vehicles, spares, rent and power. In a regional network, it also pays for extending plant into lower-density settlements where each kilometre of fibre may serve fewer accounts.

JetNet's promotions show the commercial pressure. The site offers three free months to new customers, referral rewards and an extra month for annual payment on location pages. Those offers can reduce acquisition friction and bring cash forward, but they also defer revenue that could otherwise fund plant and resilience. The public pages do not expose enough stable tariff data to calculate revenue per user, and there is no public cost breakdown.

The payment recipient offers one clue about how the business is organised. Customers using bank transfer pay Multimedia Network LLC. That company is also the AS42510 participant at UA-IX, while the route registry still names the FOP. One possible explanation is a group division between resource holding, network operation and retail contracting. Another is that records evolved at different speeds. The public evidence does not settle the allocation, so financial conclusions about the FOP alone would be speculative.

Peering can improve the economics. Traffic exchanged directly at UA-IX, DTEL-IX or Giganet may avoid a longer paid-transit path, reduce latency and keep some content reachable when an international path is impaired. Multiple transit-facing networks can improve bargaining power and continuity. Yet exchange ports, cross-connects, routers and colocation also cost money, as do backup power and staff able to manage them.

The more-specific route pattern suggests active traffic engineering. Aggregates are widely visible through Eurotele-Plus and UnderNet, while more-specifics appear through numerous adjacent networks. That can distribute traffic and shape inbound paths. It also increases the operational burden: routing filters, RPKI records and failover behaviour must remain consistent. The valid route-origin authorisations are evidence of competent address stewardship, but not a financial statement.

Rural and suburban fibre economics introduce another trade-off. Passive splitters reduce field power and can share one optical line terminal port among many users. That makes coverage more affordable. The shared tree also means that a feeder failure can affect many paying customers at once, and premium redundancy requires more fibre, optics and ports. A provider can maximise nominal coverage by building one economical tree, or spend more on protected feeders and spare equipment. Public sales pages reveal the coverage decision more readily than the resilience spending.

This is the article title's practical meaning. The local bill depends on a chain of upstream routes and field repair because no single payment purchases independence from every shared asset. The public record shows AS42510 has invested in interconnection and JetNet has invested in optical expansion. It does not show whether current prices support the spare routes, batteries, crews and capacity needed for a severe multi-site failure.

Failure travels differently depending on where it begins

The same customer symptom, no internet, can arise from several failures with very different blast radii.

A customer-premises power failure is the narrowest. The fibre and provider network can remain healthy while an unpowered optical terminal or router makes service unavailable in one home. JetNet explicitly tells customers to provide backup power. The mini-UPS catalogue is a practical response, but actual runtime depends on load and battery condition.

A drop or branch cut affects one property or a small group. Logical upstream diversity does nothing for it. A technician needs to locate optical loss, obtain access and repair or replace the cable. In an apartment, the fault may be in a riser or common room. In a village, it may be on a roadside span.

A feeder cut or high-level splitter failure affects many branches. This is the principal unmeasured xPON risk because the normal tree shares upstream fibre. A protected feeder or alternative optical line terminal could reduce the outage, but JetNet does not publish such a design. The number of affected users would depend on split ratios and topology, which are also unpublished.

A powered-node failure can be local or regional. A dead optical line terminal may remove an entire PON group. A failed aggregation switch can affect several terminals. A power problem at the internet edge could disable transit and exchange paths together if they share a facility. JetNet says active nodes have backup sources but gives no duration or site separation.

An upstream failure is the area where AS42510's public design looks strongest. Two transit-facing networks and three exchange ports create alternatives. A single carrier outage need not remove every path. But recovery quality depends on route convergence and spare capacity. If the backup is congested, service may remain technically reachable while video calls, cloud work and large downloads degrade.

A wider attack or energy failure can combine these cases. Power can disappear while crews are already repairing physical damage; traffic can move to surviving links and congest them; customer batteries can expire at different times. The World Bank assessment identifies both physical repair and backup energy as continuing sector costs. The network needs resilience across layers, not just at BGP.

The people affected are similarly varied. JetNet says it serves families, IT professionals, local businesses, educational institutions and international companies. A household may lose work, school and contact with relatives. A business may lose payment access, cloud applications or voice. A school or public-facing institution may lose an operational channel. Cloudflare's estimated population suggests the routed network matters to thousands of users, but no public customer breakdown allows a more precise count.

This layered view prevents two opposite errors. It would be wrong to call the network fragile merely because its physical map is private: current operation, multiple interconnections and passive access are meaningful positives. It would be equally wrong to call it resilient because routes are visible through many autonomous systems. The failure has to be matched to the layer that can recover it.

Public signals help, but they do not close the evidence gaps

Several external records support the picture without proving the missing physical facts. PeeringDB was updated in late June and early July 2026, which is a strong sign of current network administration. UA-IX lists a live IPv4 address for AS42510. RIPE sees full IPv4 route visibility. Cloudflare observes traffic and estimates a customer population. JetNet's site map and catalogue have 2026 modifications, and the contact, order and payment surfaces are live.

These signals make an inactive-network theory implausible. They cannot establish subscriber count, contract quality, fibre ownership, busy-hour performance or restoration time. PeeringDB fields are operator-submitted. Route collectors see paths, not ducts. A website order form can stay online even if service is unavailable at a particular address. A business-data aggregator can lag a registry change or confuse people with the same name.

The operator's own latency examples, 15 milliseconds to Poland, 30 to Germany and 115 to the United States, are also snapshots without test location, date, destination, packet size or distribution. They suggest a sensible European routing position. They do not prove worst-case performance or availability. The statement that every channel switches without quality loss is an assertion until failover measurements are published.

The signals are therefore appropriate for corroboration and hypothesis. They are not substitutes for route-level, facility-level and service-level evidence.

What would raise the network evidence grade

The most valuable disclosure would be a current responsibility statement. It would identify which legal party contracts retail service, holds AS42510, owns access plant, contracts exchange and transit services, employs field staff and carries outage obligations. It need not expose personal data or sensitive site details. A plain allocation among the FOP, Multimedia Network LLC, Jet Net Telecom LLC and any other JetNet company would let customers and counterparties understand who controls each layer.

The second would be a resilience map at an appropriate level of abstraction. It could show aggregation zones, independent metro paths, optical line terminal sites and common facilities without publishing exact cable coordinates. It should distinguish owned fibre from leased strands and capacity. For each regional cluster, it should state whether there is one feeder, a ring or dual-homed access.

The third would be interconnection evidence that joins logical and physical diversity. AS42510 already has visible route breadth. A stronger account would say whether Eurotele-Plus and UnderNet enter through separate facilities, whether the three exchange ports are reached over independent transport, and which edge routers and power systems terminate them. A dated failover test should record convergence time, packet loss and peak utilisation on the surviving path.

The fourth would be a power table. For each class of active node, the operator could publish minimum tested autonomy, battery replacement policy, generator availability and recharge assumptions. It should separate customer equipment, optical line terminals, aggregation, edge routers and support systems. A network-wide statement of independence from electricity is less useful than a conservative minimum across required nodes.

The fifth would be installed-versus-usable capacity. The operator could report how many locations have GPON and XGS-PON, normal optical split ranges, aggregate edge capacity, busy-hour utilisation and reserved failover headroom. Port speeds should be distinguished from traffic and committed transit. A claim of up to 10 Gbps would then be tied to actual service areas and backhaul.

The sixth would be repair performance. Useful measures include support hours, field-dispatch hours, crew bases, median and high-percentile restoration times, the share of faults repaired on first visit and the number of major feeder outages. A statement about critical spares and escalation to third-party plant owners would explain how the regional footprint is supported.

Finally, the network needs current IPv6 evidence. A registered and globally visible IPv6 prefix, route-origin authorisation, exchange sessions and customer tests would establish dual-stack operation. IPv4 remains operationally important, but an ISP presenting a next-generation optical upgrade should be able to show the next-generation internet protocol as well.

None of these disclosures needs to reveal exploitable detail. Aggregate zones, minimums, tested results and contractual roles would materially improve accountability while preserving security.

The appropriate conclusion is a bounded regional-ISP assessment

FOP "Makurin Stanislav Volodimirovich" is not merely a stale name in an old route record. AS42510 is current, globally visible and protected by valid route-origin authorisations. It has two transit-facing networks, three documented exchange attachments and a measurable user population. JetNet is soliciting customers, expanding or presenting GPON service across named Kyiv-area settlements, selling power equipment and preparing XGS-PON.

The regional-ISP hypothesis is therefore supported at the network and brand level. The evidence does not support the simpler claim that the FOP alone is today's complete retail and infrastructure operator. JetNet's own billing instruction and UA-IX membership point to Multimedia Network LLC, while company records show a wider group. That ambiguity is not cosmetic. It determines who owns, powers and repairs the system a customer pays for.

The resilience case is similarly divided. Logical interconnection is a strength. Passive optical access reduces the number of powered field nodes. Current operations are visible. But physical route separation, protected PON feeders, node autonomy, capacity headroom, crew depth and tested recovery are not published. The exact blast radius of a cut, node failure or prolonged blackout remains unknown.

The final grade is Medium for an active regional network and Weak for the exact entity and physical resilience claim. The public record justifies confidence that JetNet and AS42510 are operating. It does not justify confidence that every advertised route, node and field response has an independent fallback. For a household or business in Kyiv Oblast, the service still rests on a physical chain: powered equipment at both ends, an intact optical tree, working aggregation, viable upstream paths and technicians able to reach the break.