Skip to content
Broker news. Connected. RSS feed
FXBROKERS

A clearer view of the broker industry.

Industry ProvidersProvider profile

Solace: event distribution, messaging infrastructure and agent integration

Solace supplies the messaging layer that connects applications, cloud services and AI agents. Its event brokers run as managed cloud services, customer-operated software or appliances; financial firms can use them to distribute market, order and operational events across their own systems.

By FXBrokers.ai ResearchOperating status: ActivePublished Sources checked

Services and role in a brokerage architecture

Sections ↑

Solace Event Broker carries events between applications and connects brokers into an event mesh spanning clouds and data centres. Hierarchical topics, queues and routing determine which consumers receive each publication. In this product name, “broker” means messaging middleware: an order event can travel through it, while the financial application remains responsible for its trading workflow.[1]

Specifications01

Deployment and product roles

Deployment and product roles
ChoiceOperating modelMaterial scope
Solace CloudManaged messaging service.Cloud, software and appliance brokers can participate in the same event mesh.Source 1
Self-managed softwareCustomer operates containers, virtual machines or other supported infrastructure.Standard is free but unsupported and expressly non-production; Enterprise is the production edition with disaster recovery.Source 3
Event PortalDefine shared events, schemas and publish/subscribe operations.Export versioned AsyncAPI specifications for application developers.Source 6
Agent MeshAgents invoke configured tools and external systems.Authority follows deployed identities, scopes and tool configuration.Source 19

Guaranteed messaging uses a message spool and durable queues to retain accepted work across restarts and slow consumers. Exclusive queues use one active consumer; non-exclusive queues distribute work among consumers. Capacity still matters: an exhausted spool or another resource limit can prevent acceptance, in which case the publisher does not receive an acknowledgment.[5]

APIs, protocols and event definitions

Sections ↑
Specifications02

Integration surfaces

Integration surfaces
SurfaceDocumented mechanism
Messaging clientsC, Java/JCSMP, C#/.NET, Go, Python, JavaScript, Node.js and other client libraries include samples and developer documentation.Source 4
Open protocolsAMQP, MQTT, JMS, HTTP messaging and WebSocket serve different application environments. REST applications can use native HTTP clients.Source 4
Micro-integrationsConnectors can transform payloads, enrich data and change headers; deploy packaged integrations or develop components using a chosen framework.Source 4

Event Portal’s event APIs are design artefacts with selected event versions and associated schemas. JSON or YAML exports can include service-plan, server and binding details. Only Draft versions can be edited directly; released definitions are normally evolved through a new version. Export naming strategies resolve name collisions but do not automatically rewrite references inside schema content, so renamed schema references need attention.[6]

SEMP provides management APIs with authentication on every request. OAuth bearer tokens and management-user credentials are supported. Permissions separate global resources from Message VPN resources; a global administrator can modify all resources, while narrower roles can manage an assigned VPN. Read and write requests are also checked against the attributes being accessed.[16]

Pricing, commitments and renewal

Sections ↑

The current pricing page separates platform package, event-broker capacity and count, application services, and the services subscription. Foundation, Accelerate and Scale describe package choices. Enterprise prices require a quotation. Fixed subscriptions and annual drawdown licensing are available; micro-integrations are priced per integration and can include multiple flows.[7]

Pricing & charges03

Commercial dimensions

Commercial dimensions
ItemPublished treatment
Cloud capacityENT-250 through ENT-100K tiers vary connection, queue and storage limits; default managed configurations are HA, with standalone options.Source 7
Network transfersThe Cloud Service Schedule waives charges for the first 2 TB/month of transfers within or out of a Solace dedicated region. Excess uses the cloud provider’s list price unless additional transfer capacity was bought in advance.Source 9
Capacity-unit commitmentUsage is metered hourly and consumption starts when a service is activated. Excess commitment is billed at list price; unused units expire at term end.Source 10
Changing a committed mixServices can be reduced or exchanged without reducing purchased units. Production units can become non-production units; the reverse needs Solace agreement. Paid units are generally non-refundable.Source 10

Under the MSA, invoices are normally issued in advance in USD, with undisputed amounts due within 30 days of the invoice date; customer taxes are additional. Renewal increases may equal the greater of 5% or the preceding twelve-month US CPI increase. Subscriptions ordinarily renew for one year unless notice arrives at least 30 days before expiry, subject to the order.[8]

Implementation, recovery and evaluation

Sections ↑

The cloud schedule distinguishes a Solace-controlled cloud account or dedicated VPC from a Customer Private Region running in customer-controlled Kubernetes infrastructure. On-premises use brings customer responsibilities for the operating environment. The agreed region, capacity and deployment rights therefore need to match the architecture being procured.[9]

Self-managed Standard software supports up to 1,000 client connections and an 800 GB spool, but its published non-production restriction remains decisive. Enterprise supports production deployments and a larger capacity envelope. The product page lists Kubernetes, containers, virtualisation and major cloud environments; actual sizing and operations remain deployment-specific.[3]

The current direct offer advertises a free 30-day test drive.[7]

The separate Online Trial Agreement defaults to 90 calendar days unless the order or an agreed extension specifies otherwise. It limits use to non-commercial, non-production evaluation with non-production data, and does not oblige technical support. These are different scoped terms, so the offered trial and its accepted order should establish the actual period. On expiry, use must cease and software be purged or hardware returned unless a paid subscription is agreed.[11]

Process04

Recovery layers

Recovery layers
LayerDocumented behaviourImplementation implication
High availabilityA paired event broker can take over automatically from its mate.Redundancy within the HA arrangement is distinct from replication to another site.Source 13
Disaster recoveryReplication maintains data across Message VPNs at separate sites; clients reconnect to the alternate location.Configure the client reconnection path and configuration synchronisation.Source 13
Configuration rebuildCustomers manage configuration; rebuilding commonly creates a new service and reapplies settings through SEMPv2 and Cloud REST APIs.HA configuration mirroring is not an independent customer export.Source 12

The Cloud Service Schedule’s daily backup and recovery undertaking expressly excludes Customer Message Data. The technical backup guide directs message-backup requirements toward replication. A procurement specification should distinguish recovery of the service and its configuration from recovery of the actual messages carried through it.[9][12]

Support response and cloud availability

Sections ↑

The published product-support plan operates 24/7/365. Initial response means an engineer acknowledging receipt, measured from initial contact; Severity 1 and 2 clocks require a telephone call. Acknowledgment is separate from restoration.[14]

Specifications05

Published support and availability

Published support and availability
MeasureTargetScope
Severity 1 initial response30 minutes Standard/Essential; 15 minutes Advanced.Critical production incident with no known workaround.Source 14
Severity 2 / Severity 32 hours / next Business Day.Apply the plan’s severity definitions.Source 14
HA event-broker uptime99.95% monthly.Cloud SLA; account-level measurement and exclusions.Source 15
Single-node uptime99.5% monthly.Different target from HA.Source 15
Console, Event Portal, Insights, Mission Control99.95% monthly.Published cloud-platform commitment.Source 15

The SLA publishes 5% service credits from 98% to below the applicable target and 15% below 98%. Claims must be submitted in writing within two months of the downtime. Scheduled downtime and listed exclusions apply, including HA switchovers of up to 60 seconds. Falling below 98% in two of three consecutive months permits termination of the SLA on notice; that clause is scoped to the SLA.[15]

Rolling and long-term-support broker releases receive two years of Full Support. APIs receive four years from first general availability; supporting software outside Solace Cloud receives three years from its first general availability. New features can require additional licences and may have platform restrictions.[14]

Official contact routes

Commercial enquiriesWebsite
https://solace.com/contact/Source 20

Scope Requirements discussion and tailored demonstration.

General enquiriesEmail
[email protected]Source 20

Scope Published company enquiry address.

Production supportTelephone
+1 613 777 5450Source 14

Scope International support; call for Severity 1 and 2.

Support ticketsEmail
[email protected]Source 14

Scope Severity 3 issues and enhancement requests.

Application access and AI authority

Sections ↑

Messaging-client ACL profiles govern connections and publish/subscribe topics after authentication. The preconfigured Default profile is unrestricted and is assigned when another profile is absent. A further boundary matters for queue consumers: subscribe-topic ACLs are not checked when a client binds to a queue or topic endpoint. Binding to an endpoint carrying a disallowed topic can therefore bypass that topic restriction; endpoint access must be designed alongside topic rules.[17]

Management-session revocation also needs operational care. The SEMP documentation says changes to external LDAP or RADIUS user configuration do not automatically delete existing sessions; administrators must do that manually when required. OAuth requests use bearer authentication rather than these session cookies.[16]

Agent Mesh is permissive in a fresh deployment until RBAC is enabled. Enforced scopes can constrain agent and workflow invocation, including delegated and scheduled work. Platform-management and analytics HTTP enforcement additionally requires authentication; an unauthenticated development identity is not equivalent to a fully controlled production identity.[18]

Tools can execute actions in external systems. Most built-in tools inherit the agent’s invocation authority, with specific exceptions such as sending email. Custom tools can declare required scopes, checked before model exposure and execution. Built-in code runs in-process; remote Python or Go tools run in the separate Secure Tool Runtime. This is configurable execution authority, rather than a universal human-approval requirement.[19]

Company and financial-services context

Sections ↑

Solace Corporation is the supplier identified in the MSA. A conflicting provision in an order executed by both parties takes precedence; product schedules then govern their applicable offerings.[8]

Its published headquarters is 4000 Innovation Drive, third floor, Ottawa, Ontario, Canada. The commercial contact route offers an initial requirements conversation followed by a tailored demonstration.[20]

On 5 May 2026, Bridge Growth Partners announced the closing of a Solace continuation vehicle with US$790 million in commitments, co-led by Apogem Capital, Golub Capital, HSBC Asset Management and Schroders Capital. Bridge Growth Partners III participated and HOOPP rolled its existing minority position. The announcement traces Bridge Growth’s initial control investment to 2016.[21]

A November 2023 announcement described Finalto replacing legacy messaging infrastructure with Solace PubSub+ Platform. The stated deployment combined appliances and software in London, Copenhagen, New York and Singapore, intended to distribute market data and pre- and post-trade events across front, middle and back offices. This is a dated customer-project account: the announcement describes work being deployed, rather than independently establishing present completion or measured latency.[22]

Operator
Solace Corporation
Service scope
Enterprise event distribution, persistent messaging, integration and agent infrastructure.
Customer types
Banks; Capital-markets firms; Brokers; Enterprise technology teams
Delivery
SaaS; On premises; Hosted; Managed service; API
Pricing publication
Public and quoted pricing
Headquarters
Ottawa, Ontario, Canada
Platforms
Solace Event Broker; Solace Cloud; Solace Event Portal; Solace Agent Mesh
Pricing scope
Enterprise quotation and capacity commitments; free non-production Standard software and scoped trials.
Implementation
Managed cloud or customer-operated deployments, messaging clients and configuration APIs.
Support
24/7 product support with severity-specific initial-response conditions.
Permissions and regulatory scope
Messaging technology supporting financial workflows; customer trading and regulatory duties remain separate.

Provider metadata checked 21 Oct 2026Source 2Source 8Source 1Source 7Source 3Source 12Source 14Source 22Source 20

Additional information: acknowledgment, replay and exit

Sections ↑

A transport acknowledgment confirms receipt by the messaging layer; an application acknowledgment indicates processing is complete and permits removal from the queue. While a consumer remains connected, the broker waits indefinitely for its application acknowledgment. An unacknowledged message can be redelivered when the client disconnects or its flow closes. A transport-level success should therefore not be treated as confirmation that the trading or accounting operation has completed.[23]

The JavaScript guide explains that replication-group message IDs identify a particular copy on a specific queue or endpoint within a replication group. For deduplication across an entire workflow, it recommends a publisher-generated unique identifier in the payload. Applications that store processed-message IDs need to make the business update and ID storage atomic to avoid inconsistent retry outcomes.[24]

Replay is an enabled log with a storage quota, not a fixed-duration archive. The oldest entries are trimmed at 90% of its configured quota. Replay can target non-partitioned queues or topic endpoints from the oldest available entry, a date or a replication-group ID. Topic filters can exclude messages from the log while still delivering them normally; those excluded events cannot later be replayed from that log.[25]

The MSA permits order termination on 60 days’ written notice without a general prepaid-fee refund. Export requests must arrive during the term or within 30 days after the subscription ends and exclude Customer Message Data. Afterwards, retention is no longer required and deletion follows documentation, subject to law. Transition assistance is chargeable under an agreed statement of work, for at most six months after agreement termination.[8]

Cloud service deletion is irreversible and requires a Mission Control Manager or Administrator. A protected service cannot be deleted, and a service must first be removed from its event mesh. Deletion stops client access. These controls make the application cutover and independent configuration/message recovery arrangements consequential parts of an exit plan.[26]

Follow the evidence

Sources & further reading

  1. Solace Event Broker Solace · Accessed 21 Oct 2026
  2. Company Solace · Accessed 21 Oct 2026
  3. Self-Managed Software Solace · Accessed 21 Oct 2026
  4. APIs and Protocol Support Solace · Accessed 21 Oct 2026
  5. Guaranteed Messages Solace · Accessed 21 Oct 2026
  6. Event APIs Solace · Accessed 21 Oct 2026
  7. Solace Platform Pricing Solace · Accessed 21 Oct 2026
  8. Master Supply Agreement Solace · Accessed 21 Oct 2026
  9. Cloud Service Schedule Solace · Accessed 21 Oct 2026
  10. Capacity Unit Terms Solace · Accessed 21 Oct 2026
  11. Online Trial Agreement Solace · Accessed 21 Oct 2026
  12. Backing up Event Broker Services Solace · Accessed 21 Oct 2026
  13. Service Outage Protection Solace · Accessed 21 Oct 2026
  14. Technical Product Support Solace · Accessed 21 Oct 2026
  15. Cloud Service Level Agreement Solace · Accessed 21 Oct 2026
  16. SEMP Authentication and Authorization Solace · Accessed 21 Oct 2026
  17. Managing Client Access with ACL Profiles Solace · Accessed 21 Oct 2026
  18. Agent Mesh RBAC Solace · Accessed 21 Oct 2026
  19. Agent Mesh Tools Solace · Accessed 21 Oct 2026
  20. Contact Our Team Solace · Accessed 21 Oct 2026
  21. May 2026 continuation vehicle Solace · Accessed 21 Oct 2026
  22. Guaranteed Messaging Acknowledgments Solace · Accessed 21 Oct 2026
  23. Detecting Duplicate Messages Solace · Accessed 21 Oct 2026
  24. Replaying Messages in Cloud Solace · Accessed 21 Oct 2026
  25. Deleting Event Broker Services Solace · Accessed 21 Oct 2026

This profile reports dated public information, including provider-published services and terms. It does not independently test or guarantee service delivery, integrations or performance. Inclusion is not an endorsement. Permissions and protections apply only to the named legal entity, service and client category; check the current official records and your contract.

Report an error or submit an update