High severity

Rackspace Spot breach exposed Kubernetes secrets in customer clusters

Rackspace told customers that an intruder with leaked credentials could read Secrets, create pods and list nodes in affected Spot clusters, and an extended console outage slowed the clean-up.

By Mr Elite Published 7 min read

Severity
High
Status
Reported by outlets
Sources
1
Last checked

The short version

  • Rackspace said an unauthorised party used leaked credentials to access parts of Rackspace Spot on 17 November 2026, from about 00:20 to 22:00 UTC.
  • The attackers could create pods, read Kubernetes Secrets and list nodes and volumes in some customer organisations, according to the notice BleepingComputer obtained.
  • Rackspace has not said how many customers were affected or whether customer data was stolen.
  • The Spot administration console was down for an extended period; Rackspace said service was fully restored on 21 November.
  • Customers have been told to reissue tokens, fetch fresh Kubernetes configuration files and rotate every credential held in cluster Secrets.

What happened

Rackspace has warned customers that attackers broke into its Rackspace Spot cloud platform on 17 November 2026 using compromised credentials, BleepingComputer reported on 21 November. The outlet obtained the company's notification to affected customers. Rackspace Spot lets organisations deploy and manage containerised applications on Kubernetes, the widely used system for running software in containers across groups of servers.

The notification put the intrusion between about 00:20 and 22:00 UTC on 17 November. In its words, an unauthorised party "used leaked credentials to access parts of Rackspace Spot". Rackspace said it learned of the intrusion on 18 November after a customer report, and confirmed the same day that the organisation receiving the notice was affected.

According to Rackspace, the attackers could act as legitimate users in some customer organisations. That let them create pods (the small units that run containers in Kubernetes), read Kubernetes Secrets, and list nodes and volumes. Rackspace warned they could have read any Secret in an affected user's clusters, plus any Spot API tokens or refresh tokens that user held.

The company said it has revoked affected tokens and sessions, removed the unauthorised pods, rotated keys, and strengthened access controls and monitoring. It shared indicators of compromise, meaning clues that point to this attack: Kubernetes pod and object names starting with i1, glm- and probe.

Spot's administration console also went offline for an extended period, BleepingComputer reported. Customers could not manage clusters or rotate credentials that might have been exposed. Sirish Raghuram, co-founder of Platform9, which co-developed Spot with Rackspace, said on Reddit that the team chose to shut access and sanitise the whole environment. After several delays, Rackspace announced full restoration on 21 November.

Why it happened

Rackspace has said the entry point was leaked credentials. It has not said which credentials, where they leaked, or whether they belonged to customers, staff or the platform itself. BleepingComputer's report does not say.

From the notice, the sequence was this. The intruder authenticated as a legitimate user in some customer organisations. That access allowed pod creation, Secret reads, and listing of nodes and volumes. Because Secrets and Spot tokens were readable, anything stored in them was treated as exposed.

Rackspace has not confirmed that customer data was stolen. BleepingComputer asked how many customers were affected and whether data was taken, and had not received a reply when it published.

How this kind of failure works

Leaked credentials are a common way into cloud platforms. A password, API key or token ends up somewhere it should not be, such as a public code repository, a build log or a container image, and whoever finds it can log in without breaking anything. To the platform, the visitor looks like the real owner.

A Kubernetes Secret is the built-in place to keep sensitive values such as database passwords, API keys and tokens. Anyone allowed to read Secrets in a cluster can read all of them. Think of a hotel master key kept in a cabinet in the lobby: whoever opens the cabinet gets every room, not just one.

That is why one stolen login can spread. The Secrets in a cluster often hold keys to other systems, such as databases, container registries, Git repositories and cloud storage. An intruder who reads them can sign in to those systems later, from anywhere, even after the original hole is closed.

Creating pods adds another risk. A pod runs code inside the customer's environment, so an attacker can use it to look around or run their own workloads. Rackspace has not said what the unauthorised pods did, only that it removed them.

Who is affected

Rackspace has said only that some customer organisations were affected. It has not given a number. Customers using Rackspace Spot should check for the notification and look for the indicators Rackspace published.

To check, review your clusters for pods or objects whose names begin with i1, glm- or probe, and review activity from 17 November onward. A match is a strong sign of trouble, but absence of a match does not prove you were untouched, since Rackspace listed only a few indicators.

In practice, treat every credential stored in an affected cluster's Secrets as known to the attacker, along with your Spot API and refresh tokens. The risk reaches beyond Spot to any service those credentials open.

How to defend against it

Tick each step as you check it. Your progress is saved on this device.

What to learn from this

The concept: Credential blast radius

Blast radius is how much an attacker can reach with one stolen credential. A token that can read every Secret in a cluster has a wide radius, because each Secret leads to another system.

You can shrink it by giving each service its own narrowly scoped credential, setting short lifetimes so tokens expire, and keeping the ability to rotate quickly. Rackspace's own case shows the last point: customers needed the console to rotate, and it was down.

The same idea applies to cloud keys in code repositories, CI/CD pipeline tokens and shared admin accounts. In each case, the question to ask in advance is what one leak would open.

Assume any credential can leak, and limit what each one can reach so a single leak stays small.

Quick check

Sign in to earn 10 XP for a right answer

An attacker reads all Secrets in a Kubernetes cluster. Which response best limits further harm?

Timeline

  1. 17 November 2026, about 00:20 UTC Rackspace says an unauthorised party began accessing parts of Rackspace Spot using leaked credentials.
  2. 17 November 2026, about 22:00 UTC The access window described in Rackspace's notice ends.
  3. 18 November 2026 Rackspace says it learned of the intrusion from a customer report and confirmed the notice recipient was affected.
  4. After 18 November 2026 The Spot administration console goes down for an extended period while Rackspace secures the environment.
  5. 21 November 2026 Rackspace announces service is fully restored after multiple delays.
  6. 21 November 2026 BleepingComputer publishes the story based on the customer notification.

What is not known yet

  • How many customers were affected.
  • Whether customer data was stolen.
  • Which credentials leaked, and how.
  • What the unauthorised pods ran.
  • Whether the indicators Rackspace listed are complete.

Questions people ask

What happened in the Rackspace Spot breach?

Rackspace told customers that an unauthorised party used leaked credentials to access parts of its Spot platform on 17 November 2026. The attackers could create pods, read Kubernetes Secrets and list nodes and volumes in some customer organisations. BleepingComputer obtained the notification.

What are Kubernetes Secrets and why do they matter?

Secrets are the Kubernetes feature for storing sensitive values such as passwords, tokens and API keys. If an attacker reads them, they may be able to sign in to databases, cloud services and other systems. That is why Rackspace told customers to rotate them.

How do I know if I was affected by the Rackspace Spot breach?

Look for Rackspace's notification, and check your clusters for pods or objects with names starting i1, glm- or probe. Review logs from 17 November onward. Rackspace has not said how many customers were hit, so absence of indicators is not proof of safety.

What should Rackspace Spot customers do now?

Rackspace told customers to reissue Spot API tokens, replace refresh tokens in automation, download new Kubernetes configuration files and rotate credentials in cluster Secrets. It also advised reviewing logs for connected services back to 17 November.

Sources

  • BleepingComputer: Rackspace Spot breach let hackers access customer Kubernetes secrets

Know someone who should see this?

Keep reading