Skip to content
+90 850 885 27 43 info@argeict.com

ERP and accounting software

Nebim V3 and Winner infrastructure

ARGE ICT builds and operates the head-office SQL Server, store connectivity, access, and backup infrastructure for retailers running Nebim V3 or Nebim Winner, on cloud servers in ARGE Cloud or in the company's own environment.

Service architecture

The store-to-head-office data path in Nebim V3 and Winner

The POS terminal, the store server, the internet line, and the central SQL Server are links in one chain. See the path that store-to-head-office data takes.

ARGE ICT infrastructure diagram (Nebim V3 · Winner). Head office and branches connect from their own firewall to the Istanbul data center over an IPsec tunnel or MPLS/SD-WAN; remote staff use SSL VPN, ZTNA, or ARGE ZTNA. Traffic passes from the data center router through the ARGE ICT Backbone HA Cluster to an HA firewall. Inside the dedicated VPC it passes the VPC firewall and the WAF and is spread across the application servers by a load balancer; the application server (Nebim V3 · Winner), the database, and the terminal server run in this VPC; application and database tiers can be built as HA clusters on request. A service zone beside the VPC runs monitoring, SIEM and SOAR, antivirus/XDR/EDR, and DLP. It runs on Dell and Supermicro servers with Xeon Platinum and enterprise NVMe, hosts linked by a 25–200G NVIDIA Mellanox network. Backups are kept in an isolated backup zone and can optionally be replicated to a second data center (DR site). Stores connect to head office from their own firewall over IPsec, MPLS, or SD-WAN. A local database runs in each store for offline sales; mobile devices use Wi-Fi kept apart from the POS terminals. Tills and store servers at the stores, and SQL Server at head office, are highlighted.SERVICE ARCHITECTURE · NEBIM V3 AND WINNERThe store-to-head-office data path in Nebim V3 and WinnerMPLS · SD-WANInternetCustomer: head office and branchesyour own firewallData center · Istanbuldata stays in TürkiyeARGE physical environmentARGE ICT network and racksYour dedicated VPCfirewall · WAF · serversVPC monitoring and security serviceMonitoring · SIEM and SOAR · AV/XDR/EDR · DLPBackup infrastructureisolated backup zoneDR site · second data centeridentical to the primary environmentHead officeBranchesCustomer firewallIPsec tunnel endpointCustomer firewallyour company's gatewayIPsec tunnelfrom head office and branches, encryptedIPsec tunnelencrypted private linkRemote staffSSL VPN · ZTNA · ARGE ZTNARemote staffsecure access from home or fieldDC routersdata center network · HA pairDC routersterminates external linksARGE ICT Backbone HA ClusterHA link from DC network to ARGE networkARGE ICT Backbone HA Clusterredundant core backboneFirewall (HA)ARGE perimeter · HA pairFirewall (HA)only allowed traffic passesVPC firewallrules for you onlyVPC firewallentry to the VPCWAFweb application firewallWAFinspects web requestsTerminal serverRDS · desktop clientsTerminal serverremote access to the programApplication serverNebim V3 · Winner · head-office applicationHA cluster: optionalApplication serverNebim V3 · Winnerredundant cluster on requestDatabase serverCentral SQL ServerDatabase HA cluster: optionalDatabase serveryour data lives hereredundant cluster on requestInfrastructure monitoringserver, disk, network, backupInfrastructure monitoringon one dashboardSIEM and SOARlog collection and correlationSIEM and SOARalerts handled in defined stepsAntivirus · XDR · EDRendpoint protection on serversAntivirus · XDR · EDRmalware and suspicious behavior stoppedDLPdata loss preventionDLPdata egress monitoredVirtualization hostsDell and Supermicro serversXeon Platinum · enterprise NVMeVirtualization hostsenterprise-class serversfast processors and storage25G / 50G / 100G / 200G host networkNVIDIA Mellanox Ethernet cards25G / 50G / 100G / 200G host network25–200G links between serversBackup copiesseparate logical zoneseparate from productionBackup copiesyour copies are kept hereDR firewallsame rulesDR application serveractive · serving usersDR application serverstandby · takes over in a disasterDR databasesynchronized both waysDR databaseup-to-date copy, read-onlyGlobal Load Balancerhealth-checked routingGlobal Load Balancersends users to the site that is upStoresprice, promotion, and sales transferStoresevery store linked to head officePOS terminalsales and paymentStore firewallIPsec · MPLS · SD-WAN · second link advisedStore firewallstore over an encrypted linkStore databaselocal, for offline salesMobile storeAndroid · separate Wi-FiSSL VPN · ZTNAARGE ZTNAlogs and metricsBackup lineAlso to DR (active-active)traffic shared across both sitesFailover pathused only in a disasterTwo-way replicationContinuous replicationone way, primary to DR123456IPsec tunnelMPLS / SD-WANSSL VPN · ZTNAInternet to WAFDC and ARGE networkHA pairingApplication ↔ data25–200G host networkLogs and metricsBackup lineGlobal Load BalancerFailover pathDR replication1Step (below the diagram)Illustrative · HA and DR are added with the buttons on the page
IPsec tunnelMPLS / SD-WANSSL VPN · ZTNAInternet to WAFDC and ARGE networkHA pairingApplication ↔ data25–200G host networkLogs and metricsBackup lineGlobal Load BalancerFailover pathDR replicationIllustrative · HA and DR are added with the buttons on the page
Hover over or tap a component to see its name and role; choose “Show all” to see every label at once. Zoom with Ctrl/⌘ + scroll, a two-finger pinch or a double click. Use the buttons above to add HA and DR options; click a step to bring its layer forward.
The diagram shows ARGE ICT's shared infrastructure, with the parts this service touches highlighted. Application and database HA clusters and the DR site are added with the buttons above. It is illustrative; details are settled for your environment during discovery.

How to read the diagram

  1. 1
    Stores connect to head office over a redundant line

    Tills and store servers are prepared to a standard build at every store. We monitor the server, network and database side of store-to-head-office replication and spot drops and delays before users do.

  2. 2
    Traffic is filtered at the data center edge

    The connection goes from the data center router to the ARGE ICT Backbone HA Cluster and then to the firewall. All three are HA pairs side by side; if one fails, the other takes over.

  3. 3
    Nebim and SQL Server in your dedicated VPC

    The application, the terminal server (RDS) and SQL Server run in the same VPC. We size the database and schedule index and statistics maintenance; we find the cause of slowness by measuring, not guessing.

  4. 4
    Monitoring and security service beside the VPC

    Server, disk, network, and backup status show on the monitoring dashboard. Logs are collected in SIEM and SOAR, and alerts are handled in defined steps. Antivirus and XDR/EDR run on the servers, and DLP monitors unauthorized transfers of sensitive data. The scope is set in the contract.

  5. 5
    Enterprise hardware, fast internal network

    The environment runs on Dell and Supermicro servers with Xeon Platinum processors and enterprise NVMe drives. Hosts are interconnected at 25G, 50G, 100G, or 200G with NVIDIA Mellanox cards.

  6. 6
    SQL backups kept apart, restores tested

    SQL backups are copied to the isolated backup zone, and restores are tested regularly. If you want, a twin of the environment waits at the DR site.

In retail, the link between the head-office database and the stores carries sales, inventory, and campaign information. For that reason we cover this link together with the head-office SQL Server. Details on Nebim V3 ERP, Nebim V3 POS, and the store side are on the product pages.

When you need it

  • Data transfer between the stores and the head office is delayed or breaks.
  • The head-office system slows down at the start of campaigns and seasons.
  • The head-office SQL Server has a backup, but nobody has ever tested restoring it.
  • The head-office team connects to Nebim from home or from branches over an insecure path.
  • You have no backup link plan for the case where a store's link goes down.

Scope

  • Head-office SQL Server. We size and install SQL Server based on database size and user load. We schedule maintenance plans for indexes, statistics, and integrity checks.
  • Head-office and store connectivity. We monitor the server, network, and database side where the data transfer runs, so that breaks and delays are noticed early. Store register data flows to head office over these links, so link quality affects sales directly. We set up store links over an IPsec tunnel, MPLS (the carrier's private line that does not go through the internet), or SD-WAN where there is no MPLS; the store link options are compared in one place. We recommend a second link with automatic takeover for critical stores.
  • Access for head-office users. Head-office users connect to Nebim with SSL VPN, zero trust network access (ZTNA), or ARGE ZTNA. They reach only the resources they are authorized to use.
  • Peak-season capacity. Before campaigns and seasons, we review capacity, backups, and link status against the previous period's measurements. We complete any changes before the period starts.
  • Backup and disaster recovery. We copy SQL Server backups to a separate location, run restore tests, and report the result to you. We also prepare a recovery plan for losing the head-office site.
  • e-Invoice and e-Ledger connectivity. For e-Invoice, e-Archive, and e-Ledger, the head-office server must reach your e-invoicing service provider; we prepare the network and server side of that access.
  • Monitoring. We monitor the head-office server, SQL Server, backups, and store links; monitoring hours and response scope are set in the contract.

Out of scope and limits

We do not sell Nebim licenses or provide module setup or functional support, and we are not an official Nebim partner or reseller. The application settings for head-office-to-store data transfer, and which transactions a register can continue when the link is down, belong to your Nebim partner.

How we work

  1. We review the current state of the head-office SQL Server, the store links, the backups, and the access setup.
  2. We present the findings and decide link and capacity questions together, based on the store count and the link options.
  3. We build the setup and hand it over; a responsibility matrix agreed with your Nebim partner shows who does which task.
  4. After handover, we monitor connectivity, backup, and capacity status, and we share the measurements with you before peak periods.

Service models

  1. On ARGE Cloud, the head-office SQL Server runs in a virtual private cloud (VPC) dedicated to you; we build and operate it, and the stores connect to it.
  2. In your environment, you host the head-office server, and we handle setup, maintenance, store connectivity, and monitoring.
  3. As a setup that uses both, part of the head-office system runs on ARGE Cloud and part in your own server room; we draw the line together, based on data volume and store count.

Setup on ARGE Cloud

The Nebim head-office SQL Server runs on ARGE Cloud in a VPC dedicated to you; the platform is our own infrastructure in a data center in Istanbul. Store links reach the VPC over IPsec, MPLS, or SD-WAN, and a redundant firewall stands in front of it. Backups of the head-office database are kept in the isolated backup zone, separate from the VPC, and data on ARGE Cloud stays in Türkiye. Optionally, we build SQL Server as a high availability (HA) cluster and run a copy of the environment at a disaster recovery site (DR site).

Technologies we use: Microsoft SQL Server · Windows Server · IPsec · MPLS · SD-WAN · SSL VPN · ARGE ZTNA · ARGE Cloud

Frequently asked questions

Can a store keep selling if its internet link goes down? That depends on the application settings and is settled with your Nebim partner. Our part is the backup link plan for the store.

How do stores connect to the head office? They connect over IPsec, MPLS, or SD-WAN. Which one is used depends on the store count and the link options.

How do you prepare for campaigns and peak seasons? We review capacity against the measurements of the previous period. We complete the changes before the period starts.

How do you work with our Nebim partner? We work with a responsibility matrix: infrastructure is ours, and the application belongs to your partner. Who does which task is written down.

Sizing call for Nebim infrastructure

Share your store count, user count, and database size; we respond to your request within one business day.

Request a sizing call