Azure Synapse Analytics Integration
SQL pool metrics, Spark pool utilization, and pipeline execution stats for Azure Synapse Analytics. Detect DWU saturation and Spark job failures before they delay your analytics workloads.
How It Works
Create Service Principal
Register an Azure Service Principal with Monitoring Reader on your Synapse workspace resource group. TigerOps uses this to pull SQL pool, Spark pool, and pipeline metrics from Azure Monitor.
Enable Workspace Diagnostics
Turn on Diagnostic Settings for your Synapse workspace to stream SynapseRbacOperations, GatewayApiRequests, BuiltinSqlReqsEnded, and SQLSecurityAuditEvents to a Log Analytics workspace.
Enable Pool-Level Diagnostics
Separately enable Diagnostic Settings for each dedicated SQL pool and Apache Spark pool to capture query execution metrics, pool utilization, and Spark job run events.
Set DWU and Pool Alerts
Define DWU utilization thresholds for dedicated SQL pools and node utilization thresholds for Spark pools. TigerOps alerts before pool capacity is exhausted and forecasts when you need to scale.
What You Get Out of the Box
Dedicated SQL Pool Metrics
DWU utilization percentage, CPU percentage, data IO, and local tempdb percentage for dedicated SQL pools. Track query queue depth and active queries to identify pool saturation.
Spark Pool Utilization
Allocated vCores, allocated memory GB, and active applications per Spark pool. Monitor executor utilization and identify Spark jobs that overallocate cluster resources.
Pipeline Execution Stats
Synapse pipeline run counts, activity success and failure rates, and run duration percentiles. Reuses the Azure Data Factory pipeline engine — all ADF metrics apply to Synapse pipelines.
Query Performance Insights
Top queries by execution time and resource class from BuiltinSqlReqsEnded logs. Identify long-running queries, concurrency slot contention, and result set cache miss patterns.
Integration Runtime & Link
Synapse Link change feed latency, integration dataset copy throughput, and Synapse Link SQL pool lag. Monitor real-time analytics latency from OLTP source to analytics pool.
AI Workload Anomaly Detection
TigerOps AI learns normal DWU consumption and Spark job run patterns. Unusual query surges, Spark OOMKill spikes, or pipeline schedule drift are flagged with root cause context automatically.
Synapse Workspace Diagnostic Settings
Enable workspace and pool-level diagnostics for full Synapse Analytics visibility in TigerOps.
#!/bin/bash
# TigerOps — Azure Synapse Analytics diagnostic setup
SYNAPSE_WORKSPACE="my-synapse-workspace"
SQL_POOL="my-dedicated-sql-pool"
SPARK_POOL="my-spark-pool"
RESOURCE_GROUP="my-resource-group"
WORKSPACE_ID="/subscriptions/${SUBSCRIPTION_ID}/resourceGroups/${RESOURCE_GROUP}/providers/Microsoft.OperationalInsights/workspaces/tigerops-workspace"
SYNAPSE_RESOURCE="/subscriptions/${SUBSCRIPTION_ID}/resourceGroups/${RESOURCE_GROUP}/providers/Microsoft.Synapse/workspaces/${SYNAPSE_WORKSPACE}"
# Enable workspace-level diagnostics
az monitor diagnostic-settings create \
--name tigerops-synapse-workspace \
--resource "${SYNAPSE_RESOURCE}" \
--workspace "${WORKSPACE_ID}" \
--metrics '[{"category":"AllMetrics","enabled":true}]' \
--logs '[
{"category":"SynapseRbacOperations", "enabled":true},
{"category":"GatewayApiRequests", "enabled":true},
{"category":"BuiltinSqlReqsEnded", "enabled":true},
{"category":"IntegrationPipelineRuns", "enabled":true},
{"category":"IntegrationActivityRuns", "enabled":true},
{"category":"IntegrationTriggerRuns", "enabled":true}
]'
# Enable dedicated SQL pool diagnostics
az monitor diagnostic-settings create \
--name tigerops-synapse-sqlpool \
--resource "${SYNAPSE_RESOURCE}/sqlPools/${SQL_POOL}" \
--workspace "${WORKSPACE_ID}" \
--metrics '[{"category":"AllMetrics","enabled":true}]' \
--logs '[{"category":"SqlRequests","enabled":true},{"category":"DmsWorkers","enabled":true}]'
echo "Synapse Analytics diagnostics enabled for workspace: ${SYNAPSE_WORKSPACE}"Common Questions
Does TigerOps support both serverless and dedicated SQL pools?
Yes. TigerOps monitors both dedicated SQL pools with DWU-based metrics and serverless SQL pools with request counts, data processed, and query duration metrics from the BuiltinSqlReqsEnded diagnostic log.
How does TigerOps monitor Synapse Spark pools?
Spark pool metrics including allocated vCores, memory usage, and application state are collected via Azure Monitor. For deeper Spark job metrics like stage duration and shuffle bytes, TigerOps reads from the Spark event logs forwarded to the workspace Log Analytics.
Can TigerOps alert on Synapse SQL pool concurrency slot exhaustion?
Yes. TigerOps tracks the ActiveQueries and QueuedQueries metrics alongside DWU utilization. When queued queries are accumulating due to concurrency slot exhaustion, TigerOps fires an alert with the current active query breakdown by resource class.
Does TigerOps support Azure Synapse Link for Cosmos DB and Dataverse?
Yes. Synapse Link copy latency and change feed metrics are available in Azure Monitor for Synapse workspaces. TigerOps tracks these to monitor real-time analytics pipeline lag between source systems and the Synapse SQL pool.
How are Synapse Analytics alerts routed in TigerOps?
TigerOps supports routing Synapse alerts by pool name, workspace, and environment tag to different destinations including PagerDuty, Slack, Microsoft Teams, and email. You can configure separate on-call schedules for SQL pool and Spark pool incidents.
Full Observability Across Your Synapse Analytics Platform
SQL pool DWU tracking, Spark pool utilization, and pipeline monitoring for Azure Synapse Analytics. Connect in 5 minutes.