Introducing Atmosphere Releases 2.3.0, 3.4.0, and 3.4.1: Enhancing Performance, Resiliency, and Monitoring

Discover the latest Atmosphere updates with releases 2.3.0, 3.4.0, and 3.4.1, featuring enhanced performance, improved control plane resiliency, and advanced monitoring for load balancers.

We’re excited to announce the latest updates to Atmosphere with the release of versions 2.3.0, 3.4.0, and 3.4.1. These releases bring a range of critical improvements designed to enhance performance, bolster control plane resiliency, and improve monitoring and alerting for load balancer operations.
Version 2.3.0 enhances operational reliability with amphora alerting, improved metrics, and critical bug fixes. Version 3.4.0 introduces Octavia Amphora V2 for resilient load balancer failover, alongside the Valkey service, advanced alerting, and security updates. Version 3.4.1 boosts authentication performance by reactivating the Keystone auth token cache with the latest Ceph update, ensuring faster responses and a more reliable cloud environment. 

These updates reflect our continued commitment to building robust, efficient, and resilient infrastructure for modern cloud environments.  

Let us dive into the details of each release and what they mean for your operations! 

Atmosphere v2.3.0

This release focuses on enhancing monitoring, alerting, and overall system stability. Key updates include improved Octavia metric collection via the updated OpenStack database exporter, new alerting capabilities for amphoras in error states, and critical security fixes for the nginx ingress controller. Additionally, bug fixes address instance resizing issues, improve load balancer alert generation, and enhance the stability of the Cluster API driver for Magnum. 

New Features

• Octavia Metrics Collection
  The OpenStack database exporter has been updated, and Octavia metrics are now collected exclusively through it. 
• Amphora Alerting
  Added alerting mechanisms for amphoras to detect and notify when an Amphora enters an ERROR state or remains unresponsive for an extended duration.

Security Issues

• Nginx Ingress Controller Upgrade
  Upgraded from version 1.10.1 to 1.12.1 to address multiple CVEs, including:
  - CVE-2025-1097 
  - CVE-2025-1098 
  - CVE-2025-1974 
  - CVE-2025-24513 
  - CVE-2025-24514 

Bug Fixes

• Cluster API Driver Update
  Upgraded the Cluster API driver for Magnum to version 0.28.0, improving stability, introducing new features, and addressing existing bugs
• Instance Resizing Fix
  Resolved an issue where instances not booted from volumes would fail to resize due to a missing newline in SSH keys. Proper handling ensures seamless resizing for all instances. 
• Improved Load Balancer Alerts
  Enhanced alert generation for load balancers in a non-ACTIVE provisioning state while marked as ONLINE. This ensures accurate and timely alerts for transitional states like PENDING_UPDATE. 

The most important change in this release is the addition of alerting for amphoras to identify cases where an Amphora enters an ERROR state or remains unresponsive for an unexpected duration.  
 
This change significantly enhances the monitoring and operational reliability of load balancers by enabling proactive issue detection and resolution. It ensures that administrators are promptly alerted to potential problems, minimizing downtime and improving the overall stability of services reliant on load balancers.

Atmosphere v3.4.0 

This release brings significant advancements in load balancing and control plane resiliency with the introduction of Octavia Amphora V2, now enabled by default. Key updates include the addition of the Valkey service, enhanced metric collection and alerting for amphoras, critical security fixes, and multiple bug resolutions to improve overall stability and performance. 

New Features

• Valkey Service
  Introduced as a new service on Atmosphere, Valkey is required to enable Octavia Amphora V2 support. 
• Octavia Amphora V2 Support
  Now enabled by default, the Amphora V2 provider driver enhances control plane resiliency. In the event of a control plane host failure during load balancer provisioning, an alternate controller can resume and complete the operation. This resolves issues with resources stuck in PENDING_* states through persistent task state tracking and job monitoring via Jobboard. 
• Octavia Metrics Collection
  The OpenStack database exporter has been updated to handle Octavia metrics collection exclusively. 
• Amphora Alerting
  Added alerting capabilities for amphoras to detect and notify when an Amphora enters an ERROR state or remains unresponsive for an extended duration.

Security Issues

• Nginx Ingress Controller Upgrade
  Upgraded from version 1.10.1 to 1.12.1 to address multiple CVEs, including:
  - CVE-2025-1097 
  - CVE-2025-1098 
  - CVE-2025-1974 
  - CVE-2025-24513 
  - CVE-2025-24514 

Bug Fixes 

• Octavia Redis Driver
  Backported fixes to support authentication and SSL for Redis Sentinel, including multiple Sentinel server configurations. 
• Cluster API Driver Update
  Upgraded the Cluster API driver for Magnum to version 0.28.0, improving stability, fixing bugs, and adding new features. 
• Instance Resizing Issue
  Resolved an issue where instances not booted from volume would fail to resize due to a missing newline in SSH keys. Proper handling ensures smooth resizing for all instances. 
• Improved Load Balancer Alerts
  Enhanced alert generation for load balancers in non-ACTIVE provisioning states despite being marked as ONLINE. Transitional states like PENDING_UPDATE now properly trigger alerts to ensure accurate monitoring and timely responses. 

The most important change in this release is the support and default enablement of Octavia Amphora V2. This update enhances control plane resiliency by allowing alternate controllers to resume and complete in-progress load balancer provisioning during control plane host failures. By leveraging persistent task state tracking and monitoring through Jobboard, it effectively resolves the issue of resources being stuck in PENDING states.

This change is critical for ensuring the reliability and stability of load balancer operations. It minimizes downtime, reduces operational disruptions, and provides a more resilient infrastructure that can handle failures without manual intervention. This is essential for environments relying on high availability and seamless load balancing.

Atmosphere v3.4.1 

This release introduces key enhancements, including performance optimizations, critical bug fixes, improved stability for database clusters, and updated configurations to streamline authentication, storage, and orchestration services. 

New Features

• Keystone Auth Token Cache Reactivation
  The auth token cache, previously disabled due to a bug in Ceph (issue 64094), is now reactivated with the fix included in Ceph version 18.2.7. 
• Percona XtraDB Cluster Upgrade
  Updated the Percona XtraDB Cluster operator from version 1.14.0 to 1.16.1 and the cluster from 8.0.36-28.1 to 8.0.41-32.1, bringing performance improvements and bug fixes.

Upgrade Notes 

• MySQL max_allowed_packet Setting
  Increased from 4M to 16M to support larger queries. MySQL 8.x now defaults to 64M, so the configuration no longer explicitly specifies this setting

Bug Fixes

• [Cinder]/auth_type Configuration Fix
  Resolved an issue where the Cinder configuration section did not render properly. The auth_type is now set to password, ensuring full rendering. 
• Cluster API Driver Fix
  Added a custom build for OpenStack's Cluster API driver, resolving issues that blocked Magnum cluster upgrades for specific network or subnet configurations. 
• Manila Improvements:  
  - Now uses Nova micro-version 2.60, enabling support for attaching multiple volumes to an instance.
  - Connects to internal Nova and Glance endpoints instead of public ones, improving performance and reducing reliance on external network paths. 
• OAuth2 Proxy Configuration Fix
  Updated to allow API access using valid JWT tokens without interactive login, enabling programmatic access to services like Alertmanager. 
• Percona XtraDB Cluster Stability
  - Increased liveness probe timeouts (timeoutSeconds: 60, failureThreshold: 100) to prevent unnecessary restarts during prolonged operations. 
  - Changed MySQL exporter sidecar liveness checks to readiness checks, improving cluster stability during long SST operations. 
• QEMU VNC and API TLS Certificate Renewal
  Fixed an issue that prevented certificate renewal, resolving VM console access issues and live migration failures. 

Other Notes

• Database Backup and Restore Documentation
  Added new documentation to guide users on database backup and restore procedures. 

The most important change in this release is the reactivation of the Keystone auth token cache, made possible by the resolution of an upstream bug in Ceph (issue 64094) and its inclusion in Ceph version 18.2.7. 

This change significantly enhances the performance, scalability, and reliability of Atmosphere's authentication processes. Reactivating the auth token cache reduces the load on the Keystone service by minimizing database queries, leading to faster API responses and more efficient resource utilization. As authentication is a critical component of any cloud environment, this update strengthens the foundation for smoother operations and better user experiences. 

As usual, we encourage our users to follow the progress of Atmosphere to leverage the full potential of these updates.  

If you require support or are interested in trying Atmosphere, reach out to us! Our team is prepared to assist you in harnessing the power of these new features and ensuring that your cloud infrastructure remains at the forefront of innovation and reliability. Contact us today to explore how these updates can enhance your cloud experience and drive your business forward.