Summary
- DATARU OBLAKO LLC is not selling a virtual-machine line item in the narrow sense. The paid unit is a cloud virtual machine, migration support and a local continuity account: compute and storage in a Russian facility, help moving workloads from an existing stack, and recovery capacity if the old host, foreign vendor, payment path, backup chain or local server fails.
- The strongest public evidence is official and operational: Dataru's cloud page says the platform uses its own equipment in Tier III data centers, lists IaaS, S3, backup, private cloud, VDI, DRaaS, network transfer, colocation, HaaS and Kubernetes, and claims 99.95 percent availability, 24/7/365 cloud support and a familiar virtualization interface (https://dataru.ru/cloud).
- Dataru's paid account is shaped by facility and power economics, software licensing, storage and snapshot cost, backup immutability, recovery time, support labor, Russian data locality, sanctions-era vendor substitution, payment constraints and the friction of moving workloads without breaking business applications.
- Public market evidence supports the substitution thesis but not the private performance claim. Microsoft announced the suspension of new sales in Russia in March 2022 (https://blogs.microsoft.com/on-the-issues/2022/03/04/microsoft-suspends-russia-sales-ukraine-conflict/), AWS was reported as not accepting new customers in Russia and Belarus in March 2022 (https://www.axios.com/2022/03/08/russia-amazon-web-services), and Dataru's own 2025 market note frames 2022-2023 as an adaptation period after international vendor exits (https://dataru.ru/cloud/news/141/rossijskij-oblacnyj-rynok-2025-perehod-k-zrelosti-trendy-i-prognozy-ot-dataru-oblako-analitika).
- The judgement is positive but proof-bound. Dataru looks most relevant where a Russian mid-market or regional enterprise needs a practical local cloud landing zone, hands-on migration help, backup and DR options, and a supplier that can work around foreign-vendor and payment uncertainty. The judgement weakens if private data shows high support burden, poor recovery testing, weak price discipline, shallow migration capability or low customer retention against larger Russian cloud providers.
The buyer asks for a VM and prices a failed migration
The purchase begins with a small sentence from a systems manager: "We need a virtual machine." The words sound ordinary. The request may be for two vCPU, a few gigabytes of RAM, a network disk and a public or private address. A database server has outgrown a closet rack. A business application still runs on a platform the vendor no longer supports cleanly. A foreign cloud account is awkward to pay. A local server has been reliable for years, then one power event or disk failure makes the board ask why the workload had no tested place to go.
The buyer is not really buying only a VM. It is buying continuity. The cloud virtual machine is the visible unit, but the economic unit is a bundle: a compute instance that can host the application, migration support that can move it without damaging the business process, and a local continuity account that keeps the workload recoverable in Russia when the old option becomes too risky. That account includes compute and memory, disk performance, operating-system license treatment, network isolation, backup policy, recovery testing, support hours, contract currency, payment method, regulatory locality and the amount of friction the buyer can tolerate during the move.
Dataru's cloud page gives the public frame for that account. It describes business cloud services on a platform built on Dataru's own equipment in Tier III data centers, with safety, reliability, support, monitoring and resource management for business needs (https://dataru.ru/cloud). The same page lists compute tariffing, S3 storage, backup and recovery, private cloud, VDI, DRaaS, WAF, Anti-DDoS, cloud antivirus, data transfer, colocation, HaaS and Kubernetes as a Service. A buyer can read that page as a menu. The more useful interpretation is that Dataru is trying to reduce the buyer's migration risk by placing several continuity components under one commercial relationship.
That is why the first price comparison cannot be a simple ruble-per-vCPU table. A low monthly VM price is attractive only if the buyer can get the workload into the cloud, connect it to users and adjacent systems, protect it, recover it, pay for it, and operate it when staff are tired or scarce. The cost of an unsuccessful migration is not just a failed server bill. It is downtime, application rollback, a weekend of contractor labor, data inconsistency, lost confidence and another quarter trapped on an aging platform. Dataru's buyer wants to pay for less of that.
What Dataru says it sells
The official product surface is broad. On the main cloud page, Dataru presents compute and storage as the first block: virtual resources for data processing, analytics, databases and stable application operation (https://dataru.ru/cloud). It then moves quickly into adjacent services. S3 is positioned as secure, scalable cloud storage with high availability and fast file upload (https://dataru.ru/cloud/product/s3). Backup and recovery is described as a data-protection and business-continuity service with cloud storage and protected access to backup copies (https://dataru.ru/cloud/product/backup). DRaaS is positioned as replication and recovery for critical systems and data, including rapid restoration in disaster situations (https://dataru.ru/cloud/product/draas).
The private-cloud page is especially important because it makes the isolation account explicit. Dataru describes private cloud as a service based on dedicated IT infrastructure, reachable through the public internet or protected communication channels (https://dataru.ru/cloud/product/privatecloud). It says virtual machines run in a separate server cluster without sharing vCPU and RAM with other users, with network isolation and dedicated storage volumes that support individual SLA terms for IOPS and latency. It also says project activation begins with analysis of technical requirements, architecture design, virtual-host configuration, storage selection, network topology and monitoring integration.
Those claims do not prove capacity, uptime or customer satisfaction. They do show the shape of the offer. Dataru is not presenting a bare VPS shop. It is presenting a managed local cloud supplier with public cloud, private cloud, backup, disaster recovery, network transfer, colocation and administration. The administration page turns that into labor: management of virtual machines, virtual and physical network devices, 8/5 or 24/7 service levels, monitoring, vulnerability fixes, security tuning and support for both cloud and physical infrastructure (https://dataru.ru/cloud/service/admin).
The public cloud page also claims "SLA availability level 99.95 percent," "24/7/365 cloud support," a "familiar and convenient interface" for managing virtual resources, a reliable and safe environment, specialist expertise and Dataru's own verified equipment (https://dataru.ru/cloud). Those are marketing claims, but they are the right claims for a continuity product. A buyer that is migrating an accounting system, retail application, warehouse service, analytics database, internal portal or regional customer service platform does not only need capacity. It needs a supplier that can explain which parts of the stack are in the provider's responsibility, which parts remain in the customer's hands, and how recovery works when the first plan fails.
This is the first proof boundary. Public pages establish product scope and stated operating model. They do not establish whether Dataru consistently delivers 99.95 percent availability, whether its support desk shortens incidents, whether migration plans avoid rework, or whether customer renewal rates are strong. Those facts sit in private service reports, recovery tests and contract history.
The cost account starts in the facility
A VM looks software-like because it appears in a control panel, but its cost begins in a facility. Servers, storage arrays, switches, racks, cross-connects, UPS capacity, diesel backup, cooling, monitoring, physical security, spare parts and data-center rent are all inside the price. Dataru says its cloud platform is based on its own equipment in Tier III data centers (https://dataru.ru/cloud). It also says its colocation service places server equipment in advanced Moscow data centers, including Ixellerate and Datahouse, which it describes as Tier III sites built for resilience, security and scalability (https://dataru.ru/cloud).
For the buyer, that facility account matters in two directions. It lowers capital expenditure because the customer does not buy a new host, storage shelf, backup repository or spare power path. It also creates provider dependence because the customer now depends on Dataru's capacity planning, facility suppliers, power economics and hardware replacement cycle. A local cloud VM is therefore an operating-cost trade: the buyer avoids owning the facility stack, but rents Dataru's discipline in running it.
Power and facility costs are not abstract in Russia's cloud market. Growth in AI demand, GPU hosting and post-2022 local infrastructure substitution has increased attention to power density, hardware availability and data-center supply. Dataru's own December 2025 market note says Russian cloud growth was expected to slow from a 2024 boom to 28-30 percent in 2025 and 23-25 percent in 2026, while enterprise AI interest, high rates and a shortage of IT staff remained major drivers and constraints (https://dataru.ru/cloud/news/141/rossijskij-oblacnyj-rynok-2025-perehod-k-zrelosti-trendy-i-prognozy-ot-dataru-oblako-analitika). A smaller provider cannot win every scale contest against Yandex Cloud, VK Cloud, Cloud.ru or Selectel, but it can win specific accounts if its facility capacity is enough, close enough and supported well enough for the customer's migration.
The VM price therefore has to include the facility risk the buyer is escaping. If the workload stays on a local server, the business owns the next failed disk, the next power event, the next expired warranty and the next urgent hardware purchase. If it moves to Dataru, those costs become a monthly service bill and a supplier risk. The correct comparison is not old server depreciation against cloud invoice. It is old server depreciation plus replacement uncertainty, backup gaps, recovery testing, staff time, software compatibility and downtime against Dataru's recurring account.
The facility also determines whether the cloud can absorb recovery scenarios. A provider that can host a production VM but cannot hold enough backup history, burst recovery capacity or network throughput during an incident is selling compute without continuity. Dataru's public material points to the right components. The private diligence question is whether capacity, power, storage and network commitments survive real recovery exercises.
Software licenses and platform familiarity change the VM price
The buyer's second cost driver is software. A VM may need Linux, Windows Server, database licenses, backup agents, antivirus, WAF, monitoring, Kubernetes, virtualization management, middleware, application support and sometimes old enterprise software that was licensed for a physical host. The visible VM line hides the licensing account. If moving the workload changes license rights, support eligibility or upgrade obligations, the cheap VM becomes expensive.
Dataru's public cloud page says pricing for virtual machines follows a pay-as-you-go model and takes account of CPU, memory, storage type, network resources and additional options (https://dataru.ru/cloud). The page's FAQ also says customers can expand or reduce IaaS configurations to optimize IT spending in real time. That sounds like standard cloud pricing, but in Russia's current market it is not only a convenience feature. It is a way to avoid large hardware purchases while the buyer is still uncertain which imported software components, operating systems or virtualization layers can remain in use.
The official import-substitution page from Dataru Consulting gives the broader group context. It lists categories such as ERP, data warehouse, ETL and BI, and shows Western products beside local alternatives such as 1C, VK Tech, Arenadata, ClickHouse, Loginom, Apache Airflow, Visiology and Superset (https://dataru.ru/consulting/product/importsubstitution). The cloud subsidiary is not the same as the consulting business, and the article should not collapse them into one product. But the group-level message is relevant: Dataru's market is shaped by customers that need to replace, localize or redesign technology stacks after vendor exits and procurement disruption.
This is where platform familiarity becomes a commercial feature. Dataru claims a familiar and convenient virtualization interface on its cloud page (https://dataru.ru/cloud). That phrase is not a small UX promise. Migration risk rises when a customer's administrators must learn a completely new operating model while the workload is being moved. If Dataru's interface, support practice and migration team let a Russian IT department keep familiar habits while moving the workload, the buyer pays less in retraining, mistakes and contractor time. If the interface is familiar but the underlying support model is rigid, the value falls.
Competitor pricing shows why this account has to be explicit. Yandex Compute Cloud's pricing documentation says VM cost depends on allocated compute resources, operating system and time of use, with disks and network charged separately; it also says operating system use is charged depending on license and compute resources (https://yandex.cloud/ru/docs/compute/pricing). That is a useful benchmark because it makes visible what a buyer should ask Dataru: what is included in the VM price, what is separate, which licenses are billed hourly, which are customer-provided, and what happens when the workload is stopped but disks, snapshots and images remain.
The procurement risk is not paying too much for a VM. It is missing the software and operations cost around it. A serious Dataru proposal should separate compute, storage, operating-system license, marketplace or security products, backup, recovery, public IP, network transfer, support tier and migration project work. If that separation is clear, the buyer can compare Dataru against Selectel, Yandex Cloud, Cloud.ru, VK Cloud, a local data center and an internal build. If it is not clear, the VM price becomes a trap.
Backup and recovery are not add-ons
The easiest way to underprice a migration is to move the workload and postpone backup design. That is backwards. For a buyer leaving an old host, foreign account, fragile vendor stack or one-site server room, backup and recovery are part of the paid unit from day one. The buyer is not asking, "Can I boot a VM?" The buyer is asking, "Can I recover the business state when something breaks?"
Dataru's backup page is directly relevant. It describes protecting business-critical data, including databases, OS settings and files, using cloud backup placed in Tier III data centers, with dedicated disk space, recovery of damaged or lost data, and backup options from external infrastructure into Dataru's cloud, from Dataru's cloud into external infrastructure, and inside Dataru's cloud (https://dataru.ru/cloud/product/backup). The same page says internal backup covers virtual machines hosted in Dataru's cloud. That turns the VM account into a recoverability account.
The DRaaS page makes the recovery thesis clearer. It says the service replicates and restores critical systems and data, and that when the primary infrastructure fails, customer infrastructure can be started automatically on reserve capacity with minimal downtime and minimal data loss, framed through RTO and RPO (https://dataru.ru/cloud/product/draas). Public wording is not proof of actual RTO or RPO performance, but it gives the buyer the right diligence questions: what workloads are supported, how often recovery is tested, what dependencies are excluded, how DNS and network failover work, how application consistency is maintained, and who declares a disaster.
Dataru's May 2026 customer case with Enikom Nevada Group is the strongest public continuity signal, though it is company-authored. The case says the Far Eastern FMCG distributor selected Dataru cloud backup after testing local and cloud options, focusing on granular access controls, immutability and total cost of ownership. It says a joint working group prepared a migration plan in three weeks, adjusted backup-policy templates, handled the challenge of a large first copy across distance from the Russian Far East to central Russia, used a 1 Gbit/s channel, and implemented protected backups with territorial separation (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako).
That case should be used carefully. It does not prove that every Dataru migration succeeds, that the quoted recovery tests were independently audited, or that the customer's economics apply to smaller buyers. It does show the type of problem Dataru wants to own: not just storage, but a migration plan, access control, immutability, long-distance transfer, recovery testing and continuity confidence. For the assigned article's thesis, that is exactly the market signal that matters.
The buyer's budget should therefore price backup before celebrating cloud elasticity. Snapshot retention, immutable backup storage, recovery testing, data-transfer windows, backup software licensing, offsite copies, protected credentials and staff response time can outweigh the base VM. A cloud without a credible recovery practice is only another place to fail.
Support labor is the scarce unit
Local cloud substitution looks like infrastructure, but the scarce resource is often people. Russian companies that moved quickly after foreign vendor exits did not only need servers. They needed administrators, network engineers, backup specialists, database staff, security reviewers, migration planners and people who could communicate calmly during a weekend change window. Dataru's market note says the shortage of qualified IT specialists remained one of the barriers to Russian cloud growth, while that same shortage makes clouds attractive because infrastructure and support can be outsourced (https://dataru.ru/cloud/news/141/rossijskij-oblacnyj-rynok-2025-perehod-k-zrelosti-trendy-i-prognozy-ot-dataru-oblako-analitika).
Dataru's administration page sells directly into that shortage. It offers management of virtual network devices, physical network devices, virtual machines under 8/5 or 24/7 service levels, monitoring, updates, security protection and incident correction (https://dataru.ru/cloud/service/admin). The main cloud page likewise highlights 24/7/365 support (https://dataru.ru/cloud). For a buyer with a small IT team, that can be more valuable than another discount on CPU. If the customer's own staff are already maintaining ERP, warehouse systems, user endpoints, security tools and business reporting, cloud operations become a capacity problem.
The support account has two different values. The first is routine operations: patching, monitoring, access changes, capacity adjustments, backup checks, network configuration and help with normal incidents. The second is migration and recovery labor: planning cutover, copying data, validating application state, reversing a failed move, rebuilding after corruption, and coordinating with the customer's application vendor. The second is harder to price but more important when the buyer is moving away from risk.
The buyer should be suspicious of support claims that are not tied to response, responsibility and escalation. "24/7" can mean a phone is answered. It can also mean engineers with authority can act. Dataru's public pages do not publish detailed support terms, severity definitions, credits or incident metrics. That is not unusual for a custom business cloud supplier, but it means the contract and service history carry more weight than the marketing page. The buyer should ask what happens at 02:00 when a migrated VM is running but the application is unavailable because the storage path, license server, DNS record or firewall rule is wrong.
Support labor also affects retention. If Dataru becomes the team that knows the customer's workload, backup policy, network routes and recovery plan, renewal becomes easier because replacement requires rebuilding that memory. If Dataru behaves like a ticket desk that waits for the customer to diagnose everything first, the renewal weakens. The VM's economic moat is not the hypervisor. It is the operating memory around the migrated workload.
Locality is an economic feature, not only a legal checkbox
Data locality is often discussed as compliance, but in this account it is also economics. A Russian company may need personal-data compliance, domestic contractual jurisdiction, local support, ruble payment, predictable latency to Russian users, access to local data centers, and confidence that a foreign supplier will not become unreachable during a sanctions or payment disruption. The practical question is not whether locality is morally good or technically perfect. It is whether local hosting reduces uncertainty enough to justify the migration cost.
Dataru's legal and compliance surface gives the buyer several public anchors. The cloud certificate page lists documents such as GIS conformity, 152-FZ personal-data conformity, data-transmission and telematics communication licenses, MSSP certification, ISO 27001:2022 and ISO 9001:2015 certificates (https://dataru.ru/cloud/certificates). The privacy policy identifies OOO Dataru Oblako as the operator for the cloud site's personal-data processing, gives OGRN 11247700141204, and lists the legal and actual address at Paveletskaya Square 2, building 1, Moscow (https://dataru.ru/politicscloud). The contacts page gives the same Moscow business address and cloud sales/support contacts (https://dataru.ru/cloud/contacts).
Those facts do not turn every workload into a compliant deployment. Compliance depends on the customer's data, system classification, architecture, roles, controls and contract. But they do show that Dataru is selling into a market where personal-data treatment, communications licensing and Russian locality are part of the service value. Selectel's public cloud page shows the same competitive pressure: it markets cloud servers with 152-FZ compliance, multiple Russian regions and data centers, Tier III sites and personal-data protected zones up to UZ-1 (https://selectel.ru/services/cloud/servers/). Locality is a normal buying criterion in the Russian cloud market, not a Dataru-only claim.
Locality changes the buyer's risk model. If the workload handles customer records, delivery addresses, payment-adjacent data, employee information, logistics data or regulated public-sector data, the cost of a nonlocal or ambiguous cloud can include legal review, cross-border-transfer analysis, extra controls, management hesitation and a future forced migration. A local provider can reduce that uncertainty. The price is provider concentration: the buyer must trust that the local provider can keep up with security, capacity, patching and recovery better than the old internal environment or a larger Russian alternative.
Locality also matters to recovery. Dataru's Enikom case is notable because it describes a Far Eastern customer backing up into the central part of Russia, not just keeping everything beside production (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako). That makes locality more nuanced. The buyer may want data in Russia, but not necessarily in the same city. A good continuity account uses domestic geography to create separation while staying inside the buyer's legal and operational comfort zone.
The buyer should therefore avoid treating "Russia-based" as a magic word. The proper questions are where production runs, where backups sit, who can access them, how keys are managed, what evidence supports the compliance claim, how recovery is tested, and how the provider handles a request to export or delete data. Locality is valuable when it reduces real legal, payment, latency and recovery risk. It is weak when it is only a label on a hosting invoice.
Sanctions and payment constraints raise migration value
The sanctions-era cloud problem is not limited to whether a foreign service is technically blocked. Many services continue to be reachable in some form, but the account can become hard to buy, renew, support, pay for or audit. That uncertainty changes the value of a local VM. The buyer may not be migrating because the old platform failed yesterday. It may be migrating because the old platform became strategically ambiguous.
Microsoft's March 2022 statement is a clear public anchor: it announced the suspension of all new sales of Microsoft products and services in Russia and said Microsoft was stopping many aspects of its Russian business in compliance with governmental sanctions decisions (https://blogs.microsoft.com/on-the-issues/2022/03/04/microsoft-suspends-russia-sales-ukraine-conflict/). AWS was reported shortly after as not accepting new customers in Russia and Belarus, with AWS saying it had no data centers, infrastructure or offices in Russia and that its Russian business was mostly with multinational companies with local development teams (https://www.axios.com/2022/03/08/russia-amazon-web-services). These examples do not mean every Russian business lost every foreign cloud immediately. They do show why boards began asking whether dependency on foreign software and cloud channels could survive procurement, payment and support disruption.
Dataru's own market commentary frames 2022-2023 as a period of adaptation caused by international vendor exits, security risks and insufficient maturity of local alternatives, followed by a 2024 boom in import substitution among large businesses and a 2025-2026 maturity phase led more by mid-market adoption, resilience and existing project scaling (https://dataru.ru/cloud/news/141/rossijskij-oblacnyj-rynok-2025-perehod-k-zrelosti-trendy-i-prognozy-ot-dataru-oblako-analitika). That is company-authored analysis, but it matches the buying logic: the first wave is emergency substitution; the second wave is disciplined continuity.
Payment constraints deserve separate treatment. A foreign cloud account can be technically adequate and commercially awkward. If the customer cannot use normal cards, cannot get invoices accepted cleanly, cannot receive standard vendor support, or cannot guarantee license renewal, it has operational risk before any server fails. A Russian provider with local contracts and local billing can turn a foreign-payment problem into a predictable domestic operating expense. That matters for finance teams as much as for engineers.
Migration friction is the counterweight. Moving a workload from a foreign cloud, old VMware cluster, physical server or local hosting provider is not free. Data must be copied, images converted, IP ranges and DNS changed, licenses checked, access policies rebuilt, backup jobs rewritten, monitoring replaced and users warned. If Dataru's migration support is strong, that friction becomes part of the provider's value. If it is weak, the customer is simply paying a local invoice while still bearing the hardest work.
The buyer should price sanctions risk as probability times impact, not as a slogan. The question is: what happens if the current supplier cannot renew, cannot accept payment, cannot provide support, cannot supply a license, or cannot satisfy a data-locality review? If the answer is severe downtime or another rushed move, Dataru's local VM plus support and recovery account becomes insurance. If the current workload is low-stakes and easy to rebuild, the insurance value is lower.
Competitors set the benchmark
Dataru competes in a market where larger Russian cloud providers can make strong claims. Yandex Cloud publishes broad service and pricing pages, including Compute Cloud, Object Storage, Cloud Backup, network load balancing, managed databases, Data Transfer, Kubernetes and security services (https://yandex.cloud/ru/prices). Its Compute Cloud documentation gives detailed pricing mechanics for vCPU, RAM, disks, images, snapshots, traffic, public IPs, operating-system licenses and reserved-volume discounts (https://yandex.cloud/ru/docs/compute/pricing). Selectel markets cloud servers with three regions, six availability zones, seventeen pools, network bandwidth up to 10 Gbit/s, three terabytes of included traffic, DDoS protection, 152-FZ positioning, Tier III data centers and a wide product portfolio (https://selectel.ru/services/cloud/servers/).
Those competitors do not make Dataru irrelevant. They define the questions Dataru must answer. A buyer can ask why it should choose Dataru rather than a larger cloud with more self-service documentation, more public price detail, more zones, more ecosystem services or a more visible developer community. Dataru's likely answer is not hyperscale breadth. It is hands-on migration, private-cloud design, backup and DR focus, local support, Dataru group infrastructure expertise and fit for customers that want a guided move rather than a purely self-service platform.
The price benchmark should be multi-layered. For a simple test VM, Yandex or Selectel may be easier to price because public calculators and docs expose more detail. For a business migration, the lowest published VM price may not win if migration, backup, access control, support and recovery are extra or harder to coordinate. Dataru can justify a premium if it reduces project risk. It cannot justify opacity if the customer cannot separate compute, storage, license, traffic, backup, support and migration costs.
The private-cloud benchmark is also different from the public VM benchmark. Dataru's private-cloud page says all compute resources are dedicated to each client, VMs run in a separate cluster, vCPU and RAM are not shared, storage volumes are dedicated and individual performance SLA terms are possible (https://dataru.ru/cloud/product/privatecloud). That is closer to renting a managed infrastructure island than buying commodity public cloud. It will not be the right product for every workload, but it fits customers that need predictable performance, isolation, control and a supplier-managed platform without buying the whole stack.
The backup and DR benchmark may be Dataru's better contest. Selectel and Yandex also offer backup and recovery services, but Dataru's public case material emphasizes migration planning, immutability, distance, territorial separation and customer-specific backup policy (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako). A buyer with a fragile internal backup process may care more about a provider's willingness to design and test the recovery account than about the absolute breadth of cloud-native services.
The substitute list remains broader than Russian clouds. The buyer can keep the workload on premises, buy new hardware, use colocation, use HaaS, hire a systems integrator, split across multiple Russian clouds, build a private cluster, or postpone migration until the next equipment or license deadline. Dataru wins when those substitutes leave more continuity risk than Dataru's proposal. It loses when the buyer can get equal recovery, support and locality from a larger provider or from its own team at lower total cost.
Customer and market signals are useful but weak
The Enikom Nevada Group case is valuable because it contains an actual buyer scenario rather than only product claims. It says the customer operates in the Russian Far East, supplies 17,000 retail points and covers up to 95 percent of the district, then selected Dataru cloud backup after comparing local and cloud options (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako). It names access control, immutability and total cost of ownership as selection criteria. It describes a three-week migration plan and a 1 Gbit/s channel to handle a large first copy from the Far East to central Russia.
That is a strong fit for the article's thesis: the buyer wanted backup, but the economic unit was continuity across distance, support and recovery. The case also shows a retention signal because the customer comment says the work with Dataru was not the first joint project. Still, the evidence is company-authored. It does not disclose contract size, failure rates, independent recovery-test results, customer churn, competitive price comparison, or whether the customer would choose Dataru again after a severe incident. It should be treated as a directional market signal, not independent proof of service quality.
The social and public-review trail is thin in accessible material. Dataru's main site and cloud pages link to VK, Rutube, Dzen and Telegram channels, and the contact page gives public sales, support, partner, HR and marketing addresses (https://dataru.ru/cloud/contacts). The cloud page lists partners such as ITGLOBAL.COM, Orion soft, Yandex Cloud, Kaspersky, Platformcraft, Multifactor, BaseALT, Red OS, UserGate, Cyberprotect and others in different sections (https://dataru.ru/cloud). These are market-presence signals. They show ecosystem positioning and public communication, but they do not prove customer satisfaction.
Forum and review evidence was not strong enough to carry claims in the public record used here. That absence should not be overread. Many Russian enterprise cloud decisions are private, and unhappy or satisfied customers may not post detailed public reviews about backup architecture, migration support or infrastructure continuity. But the absence means the article cannot claim a broad independent reputation. It can only say that Dataru has official customer material, public contacts, partner display and market commentary.
Dataru's December 2025 market note should be read the same way. It offers useful context about cloud-market maturity, multi-cloud strategy, resilience, DRaaS, backup and niche specialization (https://dataru.ru/cloud/news/141/rossijskij-oblacnyj-rynok-2025-perehod-k-zrelosti-trendy-i-prognozy-ot-dataru-oblako-analitika). It is also Dataru-authored and includes its own director's market forecast. The most defensible use is not to treat those percentages as neutral market sizing, but to observe what Dataru thinks the market is buying: resilience, hybrid strategies, DRaaS, backup and specialized provider expertise.
For a buyer, weak public signals make private diligence more important. Ask for references similar to the workload. Ask what failed in prior migrations. Ask which recovery tests did not meet target. Ask for anonymized incident and support metrics. A supplier that can discuss failures plainly is more valuable than one that offers only customer logos and perfect scenarios.
Network evidence proves footprint, not quality
Public network evidence is useful, but only if it is kept in its lane. The directory entry for DATARU OBLAKO LLC is at https://btw.media/en/directory/dataru-oblako-llc-ru. Public DNS checked during this review resolved dataru.ru to 81.177.208.59 and cloud.dataru.ru to 185.215.4.12. RIPEstat network-info for 81.177.208.59 maps the address to prefix 81.177.208.0/23 and AS215037, while RIPEstat's AS overview identifies AS215037 as "DATARUCLOUD DATARU OBLAKO LLC" and announced (https://stat.ripe.net/data/as-overview/data.json?resource=AS215037). RIPEstat announced-prefixes for AS215037 showed prefixes including 81.177.208.0/22, 81.177.208.0/23, 81.177.210.0/23 and 194.117.65.0/24 visible in the observation window (https://stat.ripe.net/data/announced-prefixes/data.json?resource=AS215037).
The WHOIS data for 81.177.208.59 is also relevant. RIPEstat converted the address to a prefix and returned inetnum 81.177.208.0/22, netname RU-DATARUCLOUD-20030115, country RU, organization ORG-DOL10-RIPE and DATARUCLOUD-MNT, with route objects originated by AS215037 (https://stat.ripe.net/data/whois/data.json?resource=81.177.208.59). That supports the claim that Dataru Cloud has public routing resources under its own name. It does not reveal the architecture of the cloud platform, the size of the service, customer traffic, latency, redundancy or service quality.
The cloud.dataru.ru marketing site resolved differently. RIPEstat WHOIS for 185.215.4.12 maps the address to 185.215.4.0/24, netname RU-TILDAPUBLISHING-20210412 and route origin AS57724 described as Tilda Publishing LLC. That is best read as a public website dependency, not as evidence about Dataru's cloud service. Many companies use separate website hosting or site builders while operating infrastructure elsewhere. The evidence should not be inflated into a claim about production hosting.
For the buyer, the network data raises practical questions. Does Dataru provide its own public IP space for customer workloads? What upstreams, private interconnects or protected channels are available? How are customer networks isolated? What bandwidth is included? How is DDoS handled? What happens when a customer needs point-to-point connectivity, reserved ports or replication across sites? Dataru's data-transfer page says it offers virtual and physical networks, VLAN, 10 Gbit/s physical ports, point-to-point data transfer, reserved point-to-point transfer and administration of virtual or physical network devices (https://dataru.ru/cloud/product/cloudtransferdata). That product surface fits the questions. The network records only prove that there is a public footprint to investigate.
The article should not create entities from ASNs, IPs, prefixes or datasets. AS215037, 81.177.208.0/22 and related RIPE records are evidence about Dataru's public routing surface. They are not companies, customers or relationships. They help explain dependency and footprint. They do not score reliability.
The VM price contains migration friction
Migration is the part of the cloud purchase that accounting teams often see too late. A new VM can be created quickly, but a workload is not a VM. It is data, users, credentials, database state, firewall rules, scheduled jobs, monitoring hooks, backup policies, vendor support, license files, IP allowlists, DNS, certificates, local scripts, documentation and habits. The buyer who ignores those pieces will compare Dataru against a cheap VM and miss the true cost.
Dataru's main cloud FAQ says specialists help with integration of services, network-policy setup, data migration, application updates and work with PaaS and SaaS tools (https://dataru.ru/cloud). The private-cloud page says activation begins with technical requirement analysis and architecture design, including host configuration, disk subsystem parameters, network topology and integration with monitoring and management systems (https://dataru.ru/cloud/product/privatecloud). The administration page adds virtual-machine and network-device management under different support models (https://dataru.ru/cloud/service/admin). Taken together, those pages suggest Dataru wants to be involved in migration work, not only post a control panel.
The buyer should still make migration work visible in the contract. Who inventories dependencies? Who converts images? Who verifies backup consistency before cutover? Who handles DNS TTLs, firewall changes and certificate renewal? Who tests restore? Who signs off application performance? Who owns rollback? If the answer is "the customer," then Dataru's value is mainly local infrastructure. If the answer is shared, with Dataru responsible for defined migration tasks and recovery tests, the provider is selling a higher-value continuity account.
Migration friction also changes timing. A business may delay cloud adoption because the old environment still works. That delay is rational until the next event: hardware replacement, license deadline, support expiration, audit finding, foreign-vendor disruption, office move, ransomware scare, backup failure, application upgrade or data-center contract change. Dataru's sales opportunity sits near those events. The buyer is easiest to convince when the cost of staying is no longer theoretical.
There is a risk of overselling migration. A provider can move infrastructure and still fail to move the operational model. If the customer's team does not understand billing, backup windows, access control, restore procedure and support escalation after migration, the workload has only changed rooms. Dataru's public promise should therefore be judged by post-migration clarity. The customer should leave with runbooks, access model, recovery schedule, contact paths, cost controls and a tested restore. Without those, the continuity account is incomplete.
The value of a local VM is therefore partly a function of friction removed. If Dataru can reduce discovery time, first-copy difficulty, backup redesign, network change risk, license confusion and support handoff, its VM is worth more than commodity compute. If the customer has to do all of that alone, the VM is just a hosting line.
Where Dataru wins
Dataru's best account is a customer that is large enough to suffer from downtime but not large enough to staff a full cloud platform team. A regional distributor, manufacturing group, retail operator, service company, public-sector contractor, local SaaS vendor or mid-sized enterprise may have business-critical workloads, but not enough specialists to design, migrate, back up, monitor and recover them alone. For that buyer, Dataru's bundle of IaaS, backup, DRaaS, private cloud, network transfer and administration is coherent.
The company also fits buyers that want local substitution without immediately committing to a hyperscale-style platform. Some organizations need to move a few workloads first: a database, application server, backup repository, file store, VDI pilot, internal web service or 1C-related environment. They may not want the breadth of a large cloud catalog. They want a supplier that can answer the phone, design the first landing zone, provide a local invoice, and keep the migration small enough to govern. Dataru's product set reads as a practical continuity stack for that buyer.
Another strong account is private cloud. If the customer's concern is noisy-neighbor risk, predictable storage latency, audit comfort or dedicated resources, Dataru's private-cloud claims are directly aligned. The page says resources are dedicated to each client, virtual machines run in a separate server cluster and storage volumes can carry individual IOPS and latency terms (https://dataru.ru/cloud/product/privatecloud). A buyer should verify those terms, but the commercial shape is clear: Dataru can sell isolation and managed infrastructure when shared public-cloud economics are not enough.
Backup-first customers are also a natural fit. The Enikom case shows Dataru publicly telling a story about immutable backup, distance, total cost of ownership and tested recovery rather than only VM deployment (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako). Many customers will start there because backup is easier to buy than a full production migration. If backup works and support is credible, production workloads may follow.
Dataru can also win where sanctions and payment uncertainty have made foreign stacks uncomfortable but the buyer is not ready for a full replatforming. The customer may want a local continuity target that preserves application behavior while reducing vendor, payment and data-location uncertainty. That is a defensible transition account. It is less ambitious than a full cloud-native transformation, but more valuable than doing nothing.
The supplier's challenge is to keep the account measurable. "Local cloud" is a broad story. The renewal will depend on whether Dataru reduces downtime, lowers support burden, improves backup confidence, shortens migration, simplifies payment, and gives management a clearer recovery plan. Those are the real product outcomes.
Where the account breaks
The account breaks first on price opacity. Dataru does not appear in the reviewed public material with the same detailed price mechanics that Yandex publishes for Compute Cloud or the same visible self-service depth that larger clouds use to win developers. That is not fatal, because many business cloud migrations are custom. But it makes proposal discipline essential. If Dataru cannot show the buyer how compute, RAM, storage type, snapshots, backup, traffic, public IPs, licenses, support and migration work are priced, the VM becomes hard to compare and easy to mistrust.
The account breaks second on support mismatch. A customer may buy 24/7 coverage believing it includes application-level recovery, while the provider believes it covers infrastructure response. Dataru's administration page lists 8/5 and 24/7 virtual-machine administration, monitoring, updates and security work (https://dataru.ru/cloud/service/admin). The contract must define the boundary. If the database is corrupt, is that Dataru's problem, the application vendor's problem or the customer's problem? If the recovery VM starts but the application cannot serve users, who stays on the call? A continuity product fails when responsibility is discovered only during the incident.
The account breaks third on scale. Large competitors can offer broader regions, bigger catalogs, more developer documentation and more visible ecosystems. Selectel markets three regions, six zones and seventeen pools for cloud servers, plus a large data-center footprint and many adjacent products (https://selectel.ru/services/cloud/servers/). Yandex Cloud offers a wide service portfolio and detailed documentation across compute, storage, networking, managed databases, data transfer and security (https://yandex.cloud/ru/prices). Dataru does not need to match every feature, but it must be clear where its smaller-provider focus is better than the larger-provider default.
The account breaks fourth on recovery proof. Backup pages and DRaaS claims create expectations. If backups are not immutable where promised, restore tests are rare, RTO/RPO are not measurable, or recovery requires manual improvisation, the continuity account loses its core value. The Enikom case mentions regular test restores and protected backups (https://dataru.ru/cloud/news/157/dalnevostocnyj-fmcg-distributor-enikom-nevada-grupp-vybral-oblacnoe-rezervnoe-kopirovanie-dataru-oblako). Buyers should ask for that same discipline in their own design, not assume it transfers automatically.
The account breaks fifth on retention. If customers use Dataru only as an emergency landing zone and later move to larger providers or back in-house, the business is project-heavy and renewal-light. If customers expand from backup into production, private cloud, support and DR, the business becomes more durable. Public evidence does not reveal which pattern dominates. That is one of the central private facts.
The final risk is complacency after migration. A successful move can make management relax. But continuity decays unless backups are tested, access is reviewed, cost is monitored, operating systems are patched, licenses are checked, support contacts are updated and recovery assumptions are revisited. Dataru's value has to persist after the migration project. Otherwise the buyer only bought temporary relief.
Proof gaps: economics, reliability and retention
The public record supports a clear commercial hypothesis. Dataru sells local cloud substitution to Russian buyers that need a practical way to move workloads, keep them recoverable, satisfy locality and support expectations, and reduce exposure to old servers or foreign vendor uncertainty. The public material proves product categories, contact identity, certificates and licenses listed by the company, a backup customer case, a stated 99.95 percent availability claim, public AS215037 routing evidence, and a market narrative centered on import substitution, resilience, DRaaS, backup and specialized providers.
What remains are proof gaps in economics, reliability and retention.
The economics gap is total cost. Public pages do not disclose Dataru's actual VM tariff, support price, migration fee, backup price, storage tiers, network charges, license treatment, service credits or private-cloud minimums. Dataru can still be a good choice, but the buyer must build a full cost model. It should compare Dataru against Yandex Cloud, Selectel, Cloud.ru, VK Cloud, colocation, new on-premises hardware, HaaS, and delayed migration. The model should include facility and power avoided, software licenses, backup storage, recovery testing, support labor, data transfer, payment reliability, migration labor and the cost of a failed move.
The reliability gap is measured performance. Public claims and certificates are useful, but buyers need failed-start and uptime records for their own workloads, backup success rates, restore-test evidence, incident response time, storage latency, network throughput, support escalation quality and RTO/RPO reports. Network evidence from RIPEstat confirms a public footprint for AS215037; it does not measure cloud reliability. A buyer should ask for recent incident summaries, maintenance history, restore-test outputs and references from similar workloads.
The retention gap is whether customers stay after the first migration. A cloud provider that wins rushed substitution projects can look busy without building durable value. Dataru's stronger story is that backup, DR, support, private cloud and migration assistance become an operating relationship. The buyer should ask whether customers expand after the first workload, whether backup customers move production workloads, whether support tickets fall after migration, and whether customers renew because recovery confidence improved rather than because moving again is painful.
The final judgement is therefore practical. DATARU OBLAKO LLC sells more than a VM when the buyer is facing migration risk. It sells a local continuity account priced through facility cost, software licenses, backup and recovery, support labor, data locality, sanctions-era substitution, payment constraints and migration friction. That account is valuable when it makes a workload easier to move, easier to operate locally and easier to recover after failure. It is replaceable when the same economics, reliability and retention can be proven more clearly by a larger cloud, a better-supported internal platform or a cheaper continuity design.