Legacy and integration

Support for systems nobody else will touch.

Much of our background is in aviation and rail, where equipment stays in service for decades and the original vendor or documentation is often gone. We support that equipment: we find out how they actually behave, keep them running, connect them to newer equipment and fix the faults that appear between products from different vendors. Most of those faults come down to a timeout, an MTU, a byte order or a certificate, and a packet capture shows which.

Why it matters

When two vendors each blame the other, the service stays degraded while the argument continues. A packet capture shows what is really happening and settles it.

We prove the cause before changing anything, take backups before and after, and record exactly what changed and why, so a fix doesn't create a new problem.

What it covers

Unsupported and undocumented equipment
Equipment and software still in service after the vendor has gone: keeping it running, documenting how it behaves and planning a safe replacement.
Network and security faults
Duplicate IPs, MTU and fragmentation black holes, spanning-tree events, asymmetric routing through stateful firewalls, VPN and NAT faults, 802.1X and RADIUS failures, multicast and IGMP snooping, and time-sync drift.
Identity and certificate faults
Kerberos and LDAP bind failures, a certificate chain that changed under a renewal, a service account whose password rotated, and config drift from a change nobody recorded.
Industrial protocol faults
Packet capture and dissection for Modbus, DNP3, IEC 60870-5-104, EtherNet/IP and OPC UA: polling rates and timeouts, DNP3 unsolicited responses and class 0/1/2/3 polls, Modbus register maps and byte/word order, OPC DA and DCOM versus OPC UA certificates and security policies.

How it runs

  1. Capture

    A mirror port on the conversation that is failing, so we are looking at what actually happens rather than what the manual says should.

  2. Find the cause

    Read the capture, the logs and the config diffs until the fault is named, not guessed.

  3. Fix in a window

    The change applied inside an agreed window, with a configuration backup taken before and after.

  4. Write it down and watch it

    A short note for the site file saying what changed and why, and where it helps, a monitoring rule so it is caught earlier next time.

Sample deliverable

ARTA CYBER Representative deliverable — client and site redacted.

Root-cause note — sample

Staff laptops failing 802.1X after the RADIUS certificate renewal

Environment: client environment (redacted) · Symptom: from the morning after a certificate renewal, some laptops land on the quarantine VLAN while the rest connect normally.

Cause. The RADIUS server certificate was renewed from a new intermediate CA. Laptops still on the older wireless profile trusted only the previous intermediate, so they rejected the new chain during EAP-TLS and never sent their own certificate. Capture radius-eap.pcapng shows those clients ending the TLS handshake with an unknown_ca alert straight after the server’s Certificate message; laptops on the newer profile complete it.

Fix. Pushed an updated wireless profile that trusts the issuing root and names the RADIUS server, rather than pinning one intermediate, through the existing device management with site IT’s agreement. Verified: the affected laptops authenticate and land on the staff VLAN, and the RADIUS log shows EAP-TLS success for each device that had been failing.

Caught earlier next time. Added a NOC rule to alert when EAP-TLS rejects rise above their normal level for more than 10 minutes, and a change-checklist step to test the profile against any renewed RADIUS certificate before it goes live.

What you receive

  • Root-cause statement with the evidence: captures, logs, config diffs.
  • The fix, applied inside an agreed window.
  • Before and after configuration backups.
  • A short note for the site file: what changed and why.
  • Where it fits, a monitoring rule so it is caught earlier next time.
OT Asset Simulator

When a fault is safe to reproduce, we rebuild it off-site in the OT Asset Simulator, virtual Modbus TCP or DNP3 devices standing in for yours, so we can try the fix somewhere that is not your live plant before we touch the real one.

OT Asset Simulator

Common questions

Something is degraded right now. How fast can you look?

The first move is a capture, which we can often set up remotely or talk your on-site staff through within the call. What we will not do is apply a guessed fix to a live service or process to look fast; we find the cause first.

What should we send you before you arrive?

A packet capture on the affected conversation if you can take one, the relevant config exports, and a note of anything that changed recently, including firmware, passwords and switch updates. There is a short checklist we can send you.

Will the fix break something else?

That is why we prove the cause with evidence before we change anything, take a backup before and after, and write down exactly what changed. A change we cannot explain is a change we do not make.

Tell us what you need.

Email, call or use the form below. The first call is free, and if the work isn't a fit for us, we'll tell you and recommend someone who is.

Book a scoping call

A free first call to talk through what you need.

Ask about a Site Assessment

A fixed-scope first engagement: one site, an assessment you can act on.