Expertise
Server infrastructure
Servers for SCADA, metering and supervisory control: virtualisation, redundancy, recovery after a failure.
What the solution consists of
A server site with no single point of failure: a cluster, storage, off-site copies and monitoring.
Virtualisation
A cluster of two or more nodes with shared storage: if a node fails, the machines restart on a healthy one.
Backup
Scheduled copies, off-site storage, restore checks under a procedure.
Monitoring
Availability, load, disk space, events; alerts to the control room and by e-mail.
Segmentation and access
SCADA and metering servers in a separate segment, role-based access, logging of logins and changes.
Power and cooling
UPS, dual power feeds, temperature and humidity control in the server room.
Documentation
Server records, diagrams, procedures for backup, recovery and updates.
How it works: a scenario
A cluster node fails: the virtual machines move to the healthy node, the duty engineer is alerted, and after the repair the load is rebalanced.
Two nodes and shared storage
The SCADA, metering, archive and monitoring machines are spread across the two cluster nodes, the data sits in shared storage, all indicators are normal.
Node failure
Node 1 goes down: traffic to it stops, monitoring records the failure and changes the site status.
Machines move to node 2
The cluster brings up the machines of node 1 on node 2 with the same storage — SCADA and metering keep running. The duty engineer receives an SMS and an e-mail.
Recovery and rebalancing
After the repair node 1 rejoins the cluster, some machines move back, and the load is spread across both nodes again.
What we do
- Virtualisation and fault-tolerant clusters for SCADA, metering and supervisory control
- Storage systems, RAID, shared cluster storage
- Scheduled backups, off-site storage of copies
- Restore checks under a procedure: test start-up of systems from a backup
- Monitoring of servers, networks and links with alerts
- Operating systems and updates in agreed maintenance windows
- Directory services, role-based access, event logging
- Time synchronisation, redundant DNS and DHCP services
- Server room power: UPS, dual feeds, temperature control
- Documentation of configurations and procedures for the operations team
How the work goes
Survey
A list of servers and services, system dependencies, state of backups, points of failure.
Design
Cluster architecture, storage, backup scheme, equipment specification.
Migration
Systems are moved into the cluster one at a time, in agreed windows, with verification and a rollback option.
Testing
Node failure, disk failure, restore from a backup, under a test programme with a report.
Support
Monitoring, updates, backup checks, a change log.
Typical tasks
A single SCADA server without a standby
We move it to a virtualisation cluster with backups.
Failures are discovered by phone calls
We set up monitoring with alerts to the control room.
Recovery has never been tested
We run a test restore under the procedure and record the result.
SCADA on an old server out of support
We move it to a virtual machine on a new cluster; the old server stays in reserve until acceptance.
Disks are filling up, archives are growing
We expand the storage and set up rotation of archives and backups.
No access control on the servers
A directory service, role-based access, logging of logins and changes.
What the customer receives
Server site diagram
Nodes, storage, networks, system dependencies, server addresses and roles.
Procedures
Backup, recovery, updates, monitoring and response to alerts.
Test reports
Node failure, restore from a backup, acceptance reports, server records.
Result: a fault-tolerant server site with monitoring and verified recovery.
Questions
Do the servers have to be replaced?
No, if their resources and vendor support allow. The survey shows what can stay and what should be replaced.
How often is recovery checked?
Under the procedure in the contract: a test restore of systems from a backup with a report on the result.
Can the servers be hosted in the cloud?
SCADA and telecontrol servers stay on site: control must not depend on external links. For analytics and reports a dedicated site is possible.
What does monitoring include?
Availability of servers and services, load, disks, links, events; alerts to the customer's recipients under agreed rules.
09Contacts
Let us discuss your engineering task
Describe the site and the task: survey, modernisation, integration or support. We will reply and propose a format of work.
What happens after you get in touch
- Phone
- +7 777 824 89 99 (WhatsApp)
- info@net-core.kz
Contact us directly
Call us, message us on WhatsApp or send an e-mail — tell us about the site and the task.