Skip to content

News

CVE-2026-16812 in on-prem VeloCloud Orchestrator: how to respond

Arista has disclosed a critical command injection flaw in on-prem VeloCloud Orchestrator, and CISA lists it as exploited. This article explains what to check, what to ask the supplier and how to plan the update.

  • Published on
Abstract pattern of navy and red shapes on a navy grid: the standard cover of an article about a security advisory.

This article was written automatically on 19 November 2026 from the information on this site and checked by automated rules; it was not reviewed by a person. Tell us if something is wrong

On 27 July 2026 Arista disclosed CVE-2026-16812, an OS command injection flaw in the on-prem version of Arista VeloCloud Orchestrator. It has a CVSS score of 10.0, the top of the scale, and its severity is critical. CISA, the US cybersecurity agency, lists it in its Known Exploited Vulnerabilities catalogue, which means it is known to be used in real attacks. This article helps you organise the response. For the technical details, read the PiLabs advisory for CVE-2026-16812.

What has been disclosed

OS command injection means that a program passes input to the operating system's command line without checking it well enough. Someone who can reach the flaw may be able to make the server run commands its owner never intended. Arista reports that on-prem VeloCloud Orchestrator exposes internal functionality to remote attackers. Remote means the attacker does not have to be inside your building or your office network.

These on-prem versions are affected:

  • 5.2.0 up to, but not including, 5.2.3.14
  • 6.1.0 up to, but not including, 6.1.3.4
  • 6.4.0 up to, but not including, 6.4.2.4

The Arista security advisory lists the fixed releases. Treat it as the single source for version numbers.

Why it matters to your organisation

An orchestrator is a management system. In general, such a system holds wide access to the network it manages, so a server that is taken over can become a problem for much more than one machine. Three facts together explain the urgency: the score is 10.0, the flaw can be reached remotely, and CISA lists it as exploited. Do not wait for the next routine maintenance window.

Who checks what

Give each task to a named person and agree a time for the answer.

  1. Infrastructure owner: find out whether you run an on-prem Orchestrator and which version it is. Look for forgotten test or lab installations as well.
  2. Security contact: check who can reach the management interface, from the internet and from internal networks. Review logs and administrator accounts for anything unexpected. Exploitation may have started before you heard of the flaw.
  3. Service manager: if a provider runs the system for you, ask them in writing whether you are affected and what they plan to do.
  4. Decision maker: appoint one person who is responsible, and set a date by which the work is reported as finished.

Questions for the supplier

Whether you ask Arista, your reseller or your managed service provider, these questions give you a usable answer:

  • Which of our installations are in an affected version range?
  • Which fixed release do you recommend for our version line?
  • Is there a mitigation we can apply before the update, and what does it change?
  • Does the update need downtime, and what happens to branch connectivity during it?
  • How can we check for signs that the system was already misused?
  • Will you confirm in writing when the work is done?

Plan the update

CISA's required action for its catalogue entry is to apply mitigations in line with vendor instructions, or to stop using the product if mitigations are not available. Use that as your baseline and follow the instructions in the Arista advisory. The NVD entry and the CISA catalogue entry give further background.

A sensible order of work is:

  1. Back up the configuration and take a snapshot of the system.
  2. Test the update on a copy if you have one.
  3. Agree a short maintenance window and tell the people who depend on the network.
  4. Install the fixed release named by the vendor.
  5. Check the version number afterwards and confirm that the network still behaves as expected.
  6. Record what you did, when, and who approved it.

If you cannot update at once, follow the vendor's mitigation guidance in the meantime. Limiting access to the management interface to a small set of trusted addresses is a common precaution, but confirm it against the vendor's instructions. If you find signs of misuse, change administrator credentials and ask for expert help before you return the system to normal use.

Where to get help

Pinet offers cybersecurity consulting and publishes security advisories in PiLabs. The PiLabs advisory for this issue is a summary of public sources. Pinet did not analyse or test the vulnerability.

More

More news

All news

How to choose a firewall size for a branch office

A firewall that is too small slows the branch down, and one that is too large wastes budget. This guide explains what to measure and what to ask before you choose.

Have a question about this news?

Write to us. A consultant will answer.