Cybersecurity researchers have uncovered three security weaknesses in Microsoft’s Azure Data Manufacturing unit Apache Airflow integration that, if efficiently exploited, might have allowed an attacker to achieve the power to conduct varied covert actions, together with knowledge exfiltration and malware deployment.
“Exploiting these flaws might permit attackers to achieve persistent entry as shadow directors over your entire Airflow Azure Kubernetes Service (AKS) cluster,” Palo Alto Networks Unit 42 stated in an evaluation revealed earlier this month.
The vulnerabilities, albeit labeled as low severity by Microsoft, are listed beneath –
- Misconfigured Kubernetes RBAC in Airflow cluster
- Misconfigured secret dealing with of Azure’s inside Geneva service, and
- Weak authentication for Geneva
In addition to acquiring unauthorized entry, the attacker might reap the benefits of the failings within the Geneva service to doubtlessly tamper with log knowledge or ship faux logs to keep away from elevating suspicion when creating new pods or accounts.
The preliminary entry approach entails crafting a directed acyclic graph (DAG) file and importing it to a personal GitHub repository linked to the Airflow cluster, or altering an current DAG file. The tip purpose is to launch a reverse shell to an exterior server as quickly because it’s imported.
To drag this off, the risk actor must first acquire write permissions to the storage account containing DAG information by using a compromised service principal or a shared entry signature (SAS) token for the information. Alternatively, they’ll break right into a Git repository utilizing leaked credentials.
Though the shell obtained on this method was discovered to be working underneath the context of the Airflow consumer in a Kubernetes pod with minimal permissions, additional evaluation recognized a service account with cluster-admin permissions linked to the Airflow runner pod.
This misconfiguration, coupled with the truth that the pod might be reachable over the web, meant that the attacker might obtain the Kubernetes command-line software kubectl and in the end take full management of your entire cluster by “deploying a privileged pod and breaking out onto the underlying node.”
The attacker might then leverage the foundation entry to the host digital machine (VM) to burrow deeper into the cloud surroundings, acquire unauthorized entry to Azure-managed inside sources, together with Geneva, a few of which grant write entry to storage accounts and occasion hubs.
“This implies a complicated attacker might modify a susceptible Airflow surroundings,” security researchers Ofir Balassiano and David Orlovsky stated. “For instance, an attacker might create new pods and new service accounts. They may additionally apply modifications to the cluster nodes themselves after which ship faux logs to Geneva with out elevating an alarm.”
“This challenge highlights the significance of rigorously managing service permissions to stop unauthorized entry. It additionally highlights the significance of monitoring the operations of essential third-party providers to stop such entry.”
The disclosure comes because the Datadog Safety Labs detailed a privilege escalation state of affairs in Azure Key Vault that would allow customers with the Key Vault Contributor position to learn or modify Key Vault contents, akin to API keys, passwords, authentication certificates, and Azure Storage SAS tokens.
The issue is that whereas a consumer with the Key Vault Contributor position had no direct entry to Key Vault knowledge over a key vault configured with entry insurance policies, it was found that the position did include permissions so as to add itself to Key Vault entry insurance policies and entry Key Vault knowledge, successfully bypassing the restriction.
“A coverage replace might include the power to record, view, replace and usually handle the info inside the important thing vault,” security researcher Katie Knowles stated. “This created a state of affairs the place a consumer with the Key Vault Contributor position might acquire entry to all Key Vault knowledge, regardless of having no [Role-Based Access Control] permission to handle permissions or view knowledge.”
Microsoft has since up to date its documentation to emphasise the entry coverage danger, stating: “To stop unauthorized entry and administration of your key vaults, keys, secrets and techniques, and certificates, it is important to restrict Contributor position entry to key vaults underneath the Entry Coverage permission mannequin.”
The event additionally follows the invention of a problem with Amazon Bedrock CloudTrail logging that made it tough to distinguish malicious queries from reliable ones made to massive language fashions (LLMs), thereby permitting unhealthy actors to conduct reconnaissance with out elevating any alert.
“Particularly, failed Bedrock API calls had been logged in the identical method as profitable calls, with out offering any particular error codes,” Sysdig researcher Alessandro Brucato stated.
“The shortage of error info in API responses might hinder detection efforts by producing false positives in CloudTrail logs. With out this element, security instruments might misread regular exercise as suspicious, resulting in pointless alerts and potential oversight of real threats.”
