Order-to-cash Birst reporting shows high latency on ICS 2023-1 when sourcing from ION Data Lake

ERP Infor CloudSuite
, , , , , , , ,
1

Birst dashboards for order-to-cash metrics are delayed by 6-8 hours on ICS 2023-1. Sales reports that should reflect near real-time data are showing stale information, causing issues with daily sales tracking and customer inquiries.

We’re using ION Data Lake as the connector between CloudSuite and Birst. The extraction schedule appears to be running, but data isn’t flowing through as quickly as it did on our previous version. The Birst dashboard timestamp shows the last successful refresh, but the actual order data is hours old.

Our sales team relies on these dashboards for same-day order status and revenue tracking. The current latency is unacceptable for operational decision-making. We suspect the issue is related to either the extraction schedule configuration or the Data Lake connector settings, but we’re not sure which to investigate first or if incremental loads are configured properly.

2

Check your Birst extraction schedule first. In ICS 2023-1, the default extraction interval changed from 1 hour to 4 hours. You’ll need to manually adjust it back in the ION Data Lake configuration if you need more frequent updates.

3

Your latency issue is caused by a combination of configuration problems introduced in ICS 2023-1.

Birst Extraction Schedule: ICS 2023-1 changed the default extraction interval from 1 hour to 4 hours to reduce system load. To restore near real-time reporting:

  1. Navigate to ION Desk > Data Lake > Extraction Jobs
  2. Locate your order-to-cash extraction job
  3. Change schedule from “Every 4 hours” to “Every 1 hour” or “Every 30 minutes” for critical dashboards
  4. For 500-800 daily orders, hourly extraction is optimal without performance impact

Data Lake Connector Settings: The connector timeout and connection pool settings need adjustment:

  1. Go to ION Desk > Data Lake > Connectors > [Your Birst Connector]
  2. Increase timeout from default 7200 seconds to 10800 seconds (3 hours)
  3. Verify connection pool size is at least 10 for concurrent extraction operations
  4. Enable “Fast Extraction Mode” which was introduced in 2023-1 specifically for Birst integration
  5. Set max retry attempts to 2 (prevents excessive retry delays on transient failures)

Incremental Load Configuration: This is critical for reducing extraction time:

  1. In CloudSuite, verify change tracking is enabled for order-to-cash entities:
    • Navigate to System Administration > Data Management > Change Tracking
    • Ensure these entities are tracked: SalesOrder, OrderLine, CustomerInvoice, PaymentTransaction
  2. In Data Lake configuration, set extraction mode to “Incremental” not “Full”
  3. Define the incremental key field (typically ModifiedDateTime or LastUpdateTimestamp)
  4. Set incremental window to “Last Extraction Time” rather than fixed time window
  5. Enable “Detect Deletes” option so deleted orders are reflected in Birst

Verification and Optimization: After configuration changes:

  1. Trigger a manual extraction and monitor the job log - should complete in under 15 minutes for incremental
  2. Verify Birst dashboard shows updated timestamp within your new extraction interval
  3. Check Data Lake metrics - incremental extractions should process only 50-100 records per run vs thousands in full extraction
  4. Monitor for 48 hours to ensure consistent performance

With proper incremental configuration and hourly extraction schedule, your order-to-cash dashboards should show data with maximum 1-hour latency, acceptable for operational decision-making. The key is ensuring change tracking is enabled so only modified orders are extracted, dramatically reducing processing time from hours to minutes.

4

Also check your Data Lake connector timeout settings. ICS 2023-1 introduced stricter timeout limits. If your extraction job is timing out before completion, it might be failing silently and retrying on the next scheduled run, which would explain the multi-hour delays. Look at the connector logs for timeout errors.

5

I’ve seen this exact issue. ICS 2023-1 changed how incremental loads work with Birst. You need to verify that your order-to-cash entities have proper change tracking enabled in CloudSuite. Without change tracking, the Data Lake connector can’t identify which records changed and defaults to full extraction, which takes much longer and creates the latency you’re seeing.