Cybersecurity

Rackspace Spot Breach Let Hackers Access Customer Kubernetes Secrets

Published By Shivam Sharma5 min read
Rackspace Spot breach, hackers accessed customer Kubernetes secrets

Rackspace has told customers that hackers used leaked credentials to break into its Rackspace Spot cloud platform on Tuesday, 17 November, reading Kubernetes Secrets inside customer clusters. The breach also knocked the platform’s admin console offline for days. Full service was restored on Saturday, 21 November, four days after the intrusion began. How many customers were affected, and whether any data was stolen, is still unanswered: BleepingComputer, which reported the breach on 21 November, said Rackspace had not responded to its questions on either point.

What the attackers reached inside customer clusters

Rackspace disclosed the breach in a notification sent to affected customers and obtained by BleepingComputer. The notification puts the intrusion in a window of about 00:20 to 22:00 UTC on 17 November, about 21 hours and 40 minutes of access. Rackspace said it learned of the intrusion on 18 November, after a customer reported it, and confirmed that customer’s organization was affected the same day.

Inside the platform, the attackers were able to act as legitimate users in some customer organizations. That let them create pods in Kubernetes clusters, read Kubernetes Secrets, and list nodes and volumes. Kubernetes Secrets hold the sensitive values applications need to reach other systems: passwords, authentication tokens and API keys for databases, cloud services and more. Rackspace warned the attackers could have read any Secret stored in an affected user’s clusters, along with any Spot API tokens or refresh tokens held by that user.

Rackspace shared three indicators of compromise with customers: the IP address 23.224.79.106, the range 23.224.0.0/15, and suspicious Kubernetes pod and object prefixes i1, glm- and probe. The company said it has revoked affected tokens and sessions, removed the unauthorized pods, rotated keys, and strengthened access controls and monitoring.

Four days from first access to restore

The incident ran four days from first access to full restoration, and the outage and the breach compounded each other: the admin console customers needed to rotate compromised credentials was itself offline for much of that time.

Date What happened Timing
17 November 2026 Intrusion: leaked credentials used to access parts of Spot, about 00:20–22:00 UTC Day 0
18 November 2026 Rackspace learns of the intrusion from a customer report; that customer’s organization confirmed affected 1 day later
18–21 November 2026 Spot admin console offline for days while Rackspace sanitizes the environment Reported outage period
21 November 2026 Rackspace announces full service restored, after multiple delays 4 days after first access

The shutdown was deliberate. Sirish Raghuram, co-founder and chief growth officer of Platform9, which co-developed Rackspace Spot with Rackspace, said in a Reddit discussion that the team shut all access down and sanitized the entire environment to protect everyone using Spot. “We made the decision that shutting all access down and sanitizing the entire environment is what we can do to best protect everyone using Spot,” he said. An independent check is consistent with the multi-day outage: VerifyMCP’s automated probes of the Spot endpoint at mcp.spot.rackspace.com last succeeded on 18 November and have failed for three days running.

What should Rackspace Spot customers do now?

Reissue your Spot API tokens, replace refresh tokens, download fresh Kubernetes configuration files and rotate every credential stored in your cluster Secrets. Then review logs for databases, registries, Git repositories, cloud storage and third-party services dating back to 17 November, when the intrusion began.

  1. Reissue your Spot API tokens and replace every refresh token used in automation tools right away.
  2. Download fresh Kubernetes configuration files for each of your clusters and reconnect only with the new files.
  3. Rotate every credential stored in your cluster Secrets, including database passwords, tokens and API keys.
  4. Review logs for databases, container registries, Git repositories, cloud storage and third-party services from 17 November.
  5. Search your clusters for pods or objects starting with i1, glm- or probe and report any matches to Rackspace.

What Rackspace has not said yet

Three questions remain open. First, how many customers were affected. Rackspace’s notification says the attackers acted as legitimate users “in some customer organizations”, with no count or percentage attached. Second, whether customer data was stolen. Rackspace has said the attackers could have read Secrets; no confirmed theft or exfiltration has been disclosed. Third, where the leaked credentials came from. Rackspace has not said how the credentials leaked, and no attacker has been named. In the reporting reviewed, no ransomware or leak-site claim is attached to this incident.

Three major incidents in four years

Three major confirmed incidents at Rackspace in four years, and the three entry points could hardly be more different. In early December 2022, the Play ransomware gang hit the company’s Hosted Exchange environment through the OWASSRF zero-day exploit chain, built on CVE-2022-41080 and CVE-2022-41082, cutting off email for customers; Rackspace later confirmed the Hosted Exchange service would not return. In September 2024, a zero-day in an undocumented third-party utility packaged with ScienceLogic’s SL1 monitoring software exposed customer account names and numbers, usernames, device and IP information and encrypted internal device agent credentials. Now, in November 2026, leaked credentials opened the door to Spot. The common thread, as we read it, is not one flaw but repeated exposure through the seams of a managed-cloud stack.

A separate November cloud breach at Barracuda is a different company and a different incident, and the two should not be confused. Barracuda disclosed within days that an attacker used an unretired AI-vendor key to reach about 10 of its cloud accounts over 44 hours on 1 and 2 November, with tooling matching the ARTEX framework. A later update said the key sat in a configuration file in the vendor’s GitHub repository. Rackspace has not yet said where its leaked credentials came from. The contrast is the point: one company named its root cause in days; the other has not named its own yet.

November has been a hard month for cloud identity more broadly. Microsoft’s Entra Connect Health flaw was reported the same week, exposing tenant metadata and administrator email addresses.

What to watch next

  • Whether Rackspace answers BleepingComputer’s questions on the customer count and data theft.
  • Whether a second major outlet reports the breach or Rackspace publishes a public statement.
  • Whether any customer reports credential abuse in logs dating back to 17 November.
  • Whether Rackspace discloses the source of the leaked credentials.