Solace Event Broker Long Term Release 10.26.0

A retrospective on one year of production-hardened updates, baked into a single milestone release.

The foundational layer of Solace Platform is Solace Event Broker, and long-term support releases represent a great opportunity to take a retrospective on the changes introduced throughout the year.

As critical infrastructure, some organizations have processes that won’t allow them to deploy an interim release to production.  Thus, the arrival of an LTS release represents a vital operational milestone and a standardized baseline they can align with for predictable lifecycle management.  And an opportunity to take stock of all the new functionality delivered over the past year.

This 10.26.0 milestone solidifies a year’s worth of production-hardened features into a single release that will be maintained with regular updates to form the basis of your production deployments. It tightens security, lowers operational friction, and improves observability. Here is a technical breakdown of what shipped.

Solace Office Hours Interview with PLM Paul Kondrat

Security Infrastructure and Authentication Topology

Securing enterprise messaging infrastructure usually requires a balancing act between rigorous protocol compliance and emergency operational access. For users that have organizational requirements around certain authentication mechanisms, the ability to disable the use of basic authentication in certain scenarios was added.

Disabling Basic Authentication for Web and API Access

Those managing security can now completely disable basic authentication for the Solace Element Management Protocol (SEMP) and Broker Manager interfaces.

• The Blueprint: You can enforce strict single sign-on (SSO) and OAuth-only flows across all graphical and programmatic web endpoints.
• The Fallback: Emergency Command Line Interface (CLI) access via SSH maintains basic authentication locally. This elegant “break glass” configuration satisfies rigid security compliance reviews without creating a catastrophic lockout risk if external identity management systems fail.

Hybrid Client Identity Chains

For application connections, the broker introduces the ability to chain authentication sources natively. Instead of forcing a binary choice between managing credentials entirely inside a central directory or entirely on the local broker, you can now check the internal credential store first and gracefully fall back to LDAP/Active Directory if a user record isn’t found.

This is incredibly useful for architectures that mix human-operated accounts with automated application service accounts, providing a smooth migration path during identity management consolidation.

Technical Guardrail

The access level required to generate encrypted debug packages has been moved from full administrative privileges to a global read-only role. This enables operations teams to independently gather telemetry for vendor support without gaining elevated write permissions, and the built-in encryption guarantees that data remains uncompromised during transit.

High-Availability Operations and Container Orchestration

Operating an event mesh at an enterprise scale requires a deep understanding of network topology and stateful lifecycle events. 10.26.0 targets high-availability edge cases directly.

Native Kubernetes Readiness Probes

Kubernetes compliant readiness endpoint to aid in rolling upgrades, eliminating the need to maintain a readiness probe using custom scripts.

Kubernetes can now accurately distinguish between a broker that is simply “alive” and one that has fully synchronized its message pool and is ready to accept production traffic. This ensures clean, predictable behavior during rolling platform upgrades.

The scripts were complicated because readiness is a distributed algorithm between both the primary and backup brokers… it just made a lot more sense to have the broker team maintain this readiness information and have it presented just as an HTTP endpoint.

Appliance and On-Premise Software Observability

This enables teams operating in heavily regulated or air-gapped environments to easily export deep system telemetry out to third-party APM platforms like Datadog, Dynatrace, or Splunk without making vendor-specific support concessions.

Monitoring egress ports on a Solace appliance ensures that outbound physical interface traffic flows smoothly and downstream client queues don’t back up, preventing network congestion, buffer bloat, and dropped messages. Network engineers troubleshooting performance bottlenecks can now leverage the show interface rates CLI command to monitor real-time throughput metrics per physical interface across the Network Acceleration Blade (NAB).

Real-Time Observability: Distributed Tracing for AMQP

As asynchronous patterns power complex workflows like Agentic AI orchestration and cross-platform streaming pipelines, structural blind spots become unacceptable.

The 10.26.0 release eliminates one of these significant blind spots by extending distributed tracing to AMQP message flows.

Any AMQP message ingested or egressed by the broker now automatically propagates W3C trace context headers. Protocol boundaries are no longer observability boundaries. AMQP-reliant financial middleware or legacy enterprise integrations now integrate cleanly with your OpenTelemetry pipelines, letting trace spans flow uninterrupted to your central observability backends.

Data Plane Enhancements

Architectural resilience is proven when things go wrong. This LTS release refines how the broker handles edge cases in payloads and dead-letter routing.

Dead Message Queue (DMQ) Eligibility Overrides

Historically, in order for a poisoned or expired message to be moved a Dead Message Queue a metadata flag (dmq-eligible) must be stamped onto the packet by the publishing client application. If a legacy publisher or third-party integration omitted this flag, the message would be dropped instead of moving to the dead message queue.

The broker solves this by introducing endpoint-level DMQ control. Architects can now configure queues and topic endpoints to route dead messages to a designated DMQ, regardless of the state of the DMQ Eligible flag.

Large-Payload Scale

Enterprise messaging occasionally demands handling exceptionally large messages. 10.26.0 extends native 30 MB guaranteed message support down to software event brokers operating on a lower scale. This allows distributed edge deployments to handle large objects natively without forcing architects to invest in higer scale licenses simply to gain message-buffer capacity. This is critical for many retail or manufacturing fan-out scenarios where the remote locations operate in non-HA environments and at lower volume.

The 10.26.0 LTS release is now generally available for software, hardware appliance, and cloud-managed deployments.