Scale Capacity
Solace Insights Monitors for Datadog Reference provides information to help you understand how to read and filter the information provided by the monitors offered by your Insights Datadog account.
These monitors track resource capacity utilization for your
Scale capacity monitors are available exclusively for
| Monitor Name | Monitor Type | Status | Default Severity Configured | Details |
|---|---|---|---|---|
| Delivered Unacknowledged Messages Utilization | Status | Active | Alert or Warning |
Unacknowledged delivered messages threshold breach. Consumers are holding a high number of messages without acknowledging them, occupying the Predicted Impact: If the limit is reached, the Recommended Actions: Identify consumer flows with high unacknowledged message counts in Broker Manager. Check for stalled ACK threads or slow processing in consumer applications. Disconnect stalled consumers to reclaim capacity. Upscale to a higher Contact Solace for assistance if required. |
| Egress Flow Utilization | Status | Active | Alert or Warning |
Egress flow capacity threshold breach. Consuming clients are using a large portion of available egress flows (consumer bindings to queues and topic endpoints). Predicted Impact: If the limit is reached, no new consumer bindings can be created. Reconnecting consumers will fail to bind and stop receiving messages. Messages accumulate undelivered on queues with no bound consumers. Recommended Actions: Audit active egress flows in Broker Manager and disconnect idle consumers. Review consumer window sizes and ensure timely message acknowledgment. Upscale to a higher Contact Solace for assistance if required. |
| Ingress Flow Utilization | Status | Active | Alert or Warning |
Ingress flow capacity threshold breach. Publishing clients are consuming a large portion of available ingress flows (Guaranteed message producer connections). Predicted Impact: If the limit is reached, no new Guaranteed message publisher flows can be opened. New or reconnecting publishers will be unable to publish Guaranteed messages and may receive NACKs, connection errors, or silent message drops. Recommended Actions: Audit active ingress flows in Broker Manager and disconnect unused publishers. Configure negative acknowledgment handling in publishing applications. Upscale to a higher Contact Solace for assistance if required. |
| Message Spool Disk Utilization | Status | Active | Alert or Warning |
Message spool disk usage threshold breach. Queued Guaranteed messages are consuming significant storage on the broker's disk-backed message spool. Predicted Impact: If the spool reaches 100% utilization, all Guaranteed message publishing will be rejected broker-wide. Publishers receive NACKs and no new Guaranteed messages can be accepted until spool space is freed. VPN-level spool quotas may cause silent message drops on some endpoints. Recommended Actions: Identify queues with large backlogs in Broker Manager and scale up or restart consumers to drain backlogs. Purge stale messages from non-critical queues or delete unused queues. Expand message spool size using the Cloud Console or upscale to a higher Contact Solace for assistance if required. |
| Queues and Topic Endpoints Utilization | Status | Active | Alert or Warning |
Queue and topic endpoint count threshold breach. The combined number of queues and topic endpoints across all Message VPNs is approaching the broker's limit. Predicted Impact: If the limit is reached, no new queues or topic endpoints can be created. Applications using dynamic endpoint provisioning will fail. Internal broker features (Config-Sync, DMR, replication, VPN Bridges) may fail to create required endpoints, potentially disrupting event mesh connectivity or HA replication. Recommended Actions: Audit provisioned endpoints in Broker Manager and delete unused durable endpoints. Identify endpoints with no consumers and no activity. Prefer queues over topic endpoints where possible. Upscale to a higher Contact Solace for assistance if required. |
| Spooled Message Count Utilization | Status | Active | Alert or Warning |
Spooled message count threshold breach. The total number of Guaranteed messages stored in the message spool is approaching the broker's limit. This limit is independent of disk space. Predicted Impact: If the limit is reached, all Guaranteed message publishing will be rejected Recommended Actions: Identify queues with the largest message backlogs in Broker Manager and restart or scale slow consumers. Purge messages from non-critical queues. Enable TTL (time-to-live) and set max-redelivery limits to prevent indefinite accumulation. Upscale to a higher Contact Solace for assistance if required. |
| Topic Subscription Utilization | Status | Active | Alert or Warning |
Subscription capacity threshold breach. The number of unique topic subscriptions is approaching the event broker's limit. Topic subscriptions register client and queue interest in receiving messages on specific topics. Predicted Impact: If the limit is reached, new subscriptions will be rejected and new subscriber applications will fail to initialize. Client reconnects will fail, causing infinite reconnect loops. In event mesh environments (DMR/MNR), remote subscription propagation fails, causing message routing failures across the mesh. The subscription table is shared across all Message VPNs—one Message VPN at capacity can starve all others. Recommended Actions: Audit current subscriptions in Broker Manager or Insights and identify unused subscriptions. Decommission inactive queues, topic endpoints, or clients. Remove or narrow overly broad wildcard subscriptions. Upscale to a higher Contact Solace for assistance if required. |
| Total Client Connections Utilization | Status | Active | Alert or Warning |
Total client connections threshold breach. The number of concurrent client connections across all protocols (SMF, MQTT, AMQP, REST, WebSocket) and Message VPNs is approaching the event broker's limit. This limit is shared with internal event broker features (DMR links, VPN Bridges, Kafka Bridge connectors, Config-Sync). Predicted Impact: If the limit is reached, all new client connection attempts will be refused across all protocols. Clients that disconnect cannot reconnect until capacity is freed. Network disruptions can trigger cascading reconnect storms. Internal broker features (DMR links, VPN Bridges, Kafka Bridge, Config-Sync) may fail to establish, breaking event mesh connectivity and data replication. Recommended Actions: Audit active connections in Broker Manager and disconnect unused or idle clients. Apply per-VPN max-connections limits and per-username limits in client profiles. Account for internal feature connections (DMR, Bridges) in capacity planning. Upscale to a higher Contact Solace for assistance if required. |
| Transacted Sessions Utilization | Status | Active | Alert or Warning |
Transacted session usage threshold breach. Applications using Local or XA (distributed) transactional messaging are consuming most available server-side transacted session resources. Both transaction types draw from the same broker-wide pool and only Guaranteed messages can participate in transactions. Predicted Impact: If the limit is reached, no new transacted sessions can be opened and transactional applications will fail. Transactional workflows (consume-transform-publish pipelines, exactly-once delivery) cannot start new transactions. XA "in-doubt" transactions from crashed Transaction Managers hold sessions indefinitely until manually resolved. Recommended Actions: Audit active transacted sessions in Broker Manager and identify clients holding sessions without actively committing or rolling back. Review transaction lifecycle in application code to ensure prompt commit or rollback. Set max-transacted-sessions per client profile. Upscale to a higher Contact Solace for assistance if required. |