SNMP Data Has Gaps

SNMP Data Has Gaps

Problem

A device collects expected data sometimes, but there are inconsistent gaps in the charts.

This usually happens when a device cannot reliably respond to SNMP requests within the timeout period — typically because the network between the ktranslate container and the polled device is experiencing bandwidth contention, packet loss, or high latency.

Another scenario is that the device is overloaded and cannot respond quickly enough. This often occurs when collecting OIDs from very large tables with a poll_time_sec that is too fast for the device to keep up with.

Solution

As a general rule, place your polling container as close to the monitored devices as possible to reduce the chance of UDP SNMP payloads not completing the round trip.

High-latency WAN links:

Edit snmp-base.yaml and increase the timeout_ms from the default of 5000 ms to a longer interval.

Unreliable connections:

Consider increasing retries from the default of 0. Be aware that a high retry count combined with a timeout that is too short is unlikely to help — it can actually increase load on the monitored device as it tries to respond to more requests before any of them complete within the timeout window.

Large table polling (e.g., busy load balancers):

It can take additional time for the device to gather the required data for large table responses. Set a longer timeout_ms and a longer poll_time_sec delay to accommodate this.