Cloud hosting provider Rackspace has warned customers of a security breach on its Rackspace Spot platform after attackers used compromised credentials to access customer environments. The intrusion allowed malicious actors to access sensitive Kubernetes configuration data, deploy unauthorized workloads, and view infrastructure details across affected accounts.

Rackspace Spot is an elastic cloud service designed to help organizations deploy and manage containerized software using Kubernetes. As reported by BleepingComputer, the unauthorized access occurred on November 17, 2026, spanning nearly 22 hours between roughly 00:20 and 22:00 UTC.

Unauthorized Access and Rogue Pod Deployments

According to customer notifications sent by Rackspace, the threat actors obtained leaked credentials that enabled them to masquerade as legitimate users within certain customer organizations. The company detected the compromise on November 18 after receiving an alert from a customer, confirming the breach of that organization on the same day.

Once inside, the intruders exercised privileges associated with the compromised user profiles. They deployed unauthorized pods directly into Kubernetes clusters, enumerated cluster nodes and storage volumes, and extracted Kubernetes Secrets. Kubernetes Secrets are used by container environments to store highly sensitive data, including API keys, database passwords, and authentication tokens needed to communicate with internal and external services.

The attackers were in a position to read any Secret stored within an impacted cluster, alongside Spot API access tokens and refresh tokens associated with affected accounts. The hosting provider has not publicly disclosed the exact number of compromised organizations or whether external customer data stored in those workloads was copied.

Infrastructure Takedown and Remediation

In response to the incident, Rackspace and partner Platform9—which co-developed the Spot platform—temporarily shut down the administration console. The extended console outage initially prevented administrators from managing their environments or updating cluster credentials.

In a discussion on Reddit, Sirish Raghuram, co-founder and Chief Growth Officer of Platform9, addressed customer frustrations regarding the management lockout. Raghuram stated that leadership made the deliberate decision to disable all access to sanitize the entire platform and protect customer workloads. Rackspace finalized its containment efforts and restored full platform functionality on November 21.

The provider stated that it has terminated unauthorized sessions, revoked compromised tokens, dismantled malicious pods, rotated internal keys, and deployed stricter monitoring and access controls across the infrastructure.

Indicators of Compromise

Rackspace published technical details to assist incident response teams in identifying unauthorized actions inside their clusters. Security teams should look for network connections linked to the following infrastructure:

  • IP Address: 23.224.79.106
  • IP Range: 23.224.0.0/15

In addition, investigators should inspect container activity and deployment logs for unusual objects or pods bearing these specific prefixes:

  • i1
  • glm-
  • probe

What You Should Do

Organizations operating workloads on Rackspace Spot must assume that any credentials residing in affected clusters are compromised. Security teams should immediately take the following actions:

  • Rotate all Kubernetes Secrets: Immediately change all database credentials, private encryption keys, API tokens, and service account passwords stored as Secrets within Spot clusters.
  • Reissue Spot access tokens: Revoke existing Spot API tokens and generate new ones. Ensure that automation pipelines, CI/CD systems, and external tools replace their existing refresh tokens.
  • Obtain fresh cluster configs: Download updated Kubernetes configuration files (kubeconfig) through the restored administrative dashboard to ensure stale credentials cannot be reused.
  • Inspect third-party audit logs: Review external service logs—including databases, code repositories, container registries, and cloud object storage—dating back to November 17, 2026, for unusual access originating from the disclosed IP addresses.
  • Enforce credential hygiene: Ensure that all user accounts tied to platform management use strong, unique authentication credentials and enforce multi-factor authentication where supported.