InnerWeb
Skip to main content

On-premise server room installation: FAQ for IT

The IT department asks the same questions before every installation: who supplies the machines, how the system sees the network and where the data sits. Below we have collected the answers in the order they come up at the first technical meeting. We describe an installation at the client's site in which the system runs in Docker containers on the client's infrastructure.

4
network zones in the on-premise installation pattern
Source: card
2
instances: production and twin pre-production
Source: card

Who provides what: division of responsibility

The first question is usually whether we supply the hardware. In an on-site installation the infrastructure, meaning the machines, the network and the disks, is provided by the client. The software vendor deploys the system in Docker containers and prepares two instances: a production instance and a twin pre-production instance. This split lets the IT department keep full control over the hardware, the host operating system and the patch policy.

InnerWeb is a software vendor, not a device manufacturer. In practice this means you do not have to buy servers from us for the installation, and the site works without GPS and without an external map service. It also does not require our own access control readers or replacing the client's existing system: if the plant already has an access control system, we integrate with it rather than replace it.

  • Client: machines, network, disks, host system, host patches
  • Vendor: software in Docker containers, production instance and pre-production
  • Jointly: installation schedule and the maintenance window

Which network zones the installation assumes

The second question concerns segmentation. The pattern for an on-premise installation covers four zones. Zone 0 is the plant area with BLE transmitters. Zone 1 is the users inside the plant. Zone 2 is the plant LAN, and zone 3 is the client's private cloud in which the system server runs.

For the IT department this has a practical meaning: firewall rules can be described zone by zone, and every flow has an owner. The transmitters are not data hosts, only a source of signal. Users connect to the system through the plant network or through the telecom operator, while the server stays within the client's control.

How external traffic reaches the server

Employees' phones on the GSM network and the vendor's service team do not connect to the server directly. All traffic from outside arrives only through the handover point under the client's control. It consists of a web application firewall (WAF), one published address, an outbound firewall and the end of the IPsec tunnel on the client's side.

From the IT perspective this means one surface to audit and one place to enter rules. The client can switch off the vendor's access with a single action, because the tunnel endpoint sits on their side. The details of the rules are described on a separate page about the handover point: WAF, IPsec and the outbound firewall.

Where data and backups are kept

Data and backups stay on the client's disks. Data does not leave the plant, so questions about transfers outside the country or about processing subcontractors for the server layer do not apply in the same form as in a service model.

The IT department should still decide who is responsible for retention policies and restore tests. We suggest recording this in the acceptance protocol: the backup location, the frequency and the date of the first test restore on pre-production.

What happens to the mobile app after losing connectivity

The mobile app works offline and queues data, which reaches the server once connectivity returns. The server must therefore accept batches of events that built up during an outage, and monitoring should distinguish a connectivity break from a server failure.

For IT the conclusion is simple: an alarm about missing events from one area does not necessarily mean the host has failed. It is worth setting time thresholds in the plant's monitoring tool and describing them in the documentation.

Documentation the IT department receives

Before the installation the IT department receives a description of the zones, a list of flows through the handover point, requirements for the machines and a description of the split between production and pre-production. Everything described here applies to an installation in which the client provides the infrastructure and the vendor deploys the software in containers.

After the installation the acceptance documentation includes a checklist confirming that both instances run, that backups are written to the client's disks and that the handover point passes only the agreed flows.

Video

Frequently asked questions

Does InnerWeb supply servers for the installation?

No. The infrastructure, meaning the machines, the network and the disks, is provided by the client. The vendor deploys the software in Docker containers in two instances: production and pre-production.

Does data leave the plant?

No. Data and backups stay on the client's disks. Traffic from outside passes through the handover point under the client's control.

Do the access control readers have to be replaced?

No. InnerWeb does not produce readers and does not replace the client's existing system, it integrates with it.

Does the system need GPS or an external map service?

No. The location works without GPS and without an external map service, so the server does not need to reach such services.

Let's talk about the installation in your server room

We will present the requirements and the split of responsibilities for your plant based on the installation documentation. Book a conversation with the InnerWeb team.

Let's talk