Troubleshooting "Number of poll responses sent onto the bus does not match the number of poll responses received off of the bus" Errors

Products

• Network Observability
• CA Performance Management

Issue/Introduction

This knowledge article explains the root cause of the "Number of poll responses sent onto the bus does not match the number of poll responses received off of the bus" error found in the Data Aggregator (DA) `PollSummary.log`. 

Contrary to common assumptions, this warning indicates DC-side throttling rather than a Data Aggregator memory or heap issue.

Environment

DX NetOps :: All Supported Versions

Cause

This warning occurs when the Data Collector (DC) deliberately suppresses poll requests due to device responsiveness issues in a specific poll group.

When devices within a poll group fail to respond to SNMP requests (indicated by a non-zero badPollRequestTimedOutCount),

the NetOps Portal assumes the devices are too busy to handle further polls.

This triggers a throttling mechanism where the DC holds back subsequent scheduled polls for that cycle.

• Verification: To confirm this, review the corresponding INFO log line immediately preceding the ERROR line in the PollSummary.log. Look for the following fields:
  • polledItemCount
  • pollRequestCount
  • notSentPollRequestCount
  • badPollRequestTimedOutCount

The calculation polledItemCount − pollRequestCount = notSentPollRequestCount confirms that the "missing" responses are actually requests the DC chose not to send.

This behavior is widespread if multiple poll groups or numerous DCs report similar patterns.

Resolution

Because this is a device-responsiveness symptom rather than a DA memory issue, increasing DA memory or CPU will not resolve these warnings. Instead, focus on the following Data Collector-side optimizations:

1. Review Poll Group Sizing: The poll groups triggering these logs may be oversized, leading to congestion. Consider reducing the number of devices per poll group.
2. Investigate SNMP Timeouts: Analyze the badPollRequestTimedOutCount to identify specific devices that are frequently timing out. These devices may be overloaded or experiencing network latency, necessitating a review of their SNMP polling intervals or device health.
3. Adjust Throttling Thresholds: Ensure that the poll group thresholds are appropriately configured for the performance capabilities of the monitored infrastructure.

Additional Information

According to the DX NetOps 25.4 Administration Guide (section: "Gaps in Data Appear during Throttling"), this behavior is an expected safeguard. The system is protecting the monitored devices from being overwhelmed while they are already experiencing SNMP timeouts.