This issue is specific to Juniper. Some of them have data for 1 or 2 response paths out of say 10 or so. I can't pinpoint model thus far but some are M10i at JUNOS 14.1R7.4 Build date:2016-02-24
Starting with one device (Juniper M10i). Ran on-demand report. All elements "CPU, Memory, Interfaces, etc.) of the device show data with the exception of response path. This device has several response paths configured. I Picked one response path to use for input into the Data Collector Debug. The debug output from "Sending Response" is error=REQUEST_TIMED_OUT.
I performed an snmpwalk on the RPM MIB to verify the RPM tables were responding and they were.
I checked the collector karaf.log file and see the associated device listed with sever others in something like this;
2022-05-18T20:35:00,001 | WARN | 300000-thread-1 | ThrottleCounterManager | or.common.ThrottleCounterManager 279 | 39 - com.ca.im.data-collection-manager.core.common - 20.2.9.RELEASE- 361 | | Polls not sent due to TIMEOUT: /xxx.xxx.xxx.xxx=177 /xxx.xxx.xxx.xxx=158.../.../...
Juniper response paths showing no data
Release : 21.2
MTU of the data collector was smaller then the MTU of the device replying to the SNMP request.
The MTU of the device(s) interface (1500) vs. the MTU of the Data Collector interface (1300) turned out to be an issue. Some of the bigger get response packets were too big for the Data Collector interface to handle and were dropped. Setting the Data Collector interface MTU to 1500 to match the device resolved the issue.