Devices show Reachability Metric data, but no "Average Response - Reachability" Metric data
Applicable to all current CA Performance Manager releases 2.7 and earlier.
During discovery, the success of ICMP and SNMP communications determine which Vendor Certifications (VC) are used for gathering Reachability data. There are two possible default VC's that could be selected.
How does the system determine which VC's to use?
How does each VC handle the "Average Response - Reachability" Metric data calculations?
Why doesn't the System Statistics VC have an Expression set for the Metric?
Adding an expression for the Average Response Metric in the System Statistics VC would require us to compute Avg. RTT for SNMP requests. Given the randomly distributed SNMP polling occurring on devices for other components, this Avg. RTT metric may be variable and the metric may not be as reliable.
Is this why my device has no data for this metric?
To determine this go to the devices "Polled Metric Families" tab in the Data Aggregator Inventory view, within the CA Performance Manager web UI. Look for the Reachability Metric Family and note the VC set against the device from that Metric Family. If using the default certification options it will either use the "ICMP" or "System Statistics" named VC.
There are a few possible fixes for this.
1: If the "System Statistics" VC is in use, first confirm that there are no Ping response issues between the Data Collector (DC) host managing the device and polling it, and the device itself. Do this by:
A: Check at the command line on the DC to see if a simple "ping <hostName" or "ping <IP_Address>" command returns without delay.
B: Check for Ping related polling errors or communication issues in the DC debug pages "Detailed Poll Logging" option. Gather and review this data as follows:
Detailed Poll Logging messages by IP
NOTE: When enabling this, if doing so for multiple IP's keep track of which ones were enabled. There is no listing of enabled IP addresses. The only way to disable the logging if the enabled IPs are unknown is to stop and restart the Data Collector dcmd service. If the enabled IPs are known, see instruction 8 below to disable them. If they are not disabled, they debug will continue to consume memory and resources.
1 - Open a web browser and launch this URL against the Data Aggregator (DA) host:
2 - Select the Enable/Disable link next to the "Detailed Poll Logging by IP" radio button option
2a - On the Enable Poll Debug Logging page:
- Set the IP address for the device
- Set the IP Domain to the one managing the device
- Leave filtering set to default value None
- Leave Cache set to default value 60
2b - Ensure the "Enable Poll Logging" option radio button is selected
2c - Select the Enable Logging button at the bottom
If successful a message like the following will be seen:
"Successfully enabled poll logging for <IP>"
For instance success for enabling against IP 126.96.36.199 would be:
"Successfully enabled poll logging for 188.8.131.52"
3 - Hit the back button on your browser to return to the "Enable Poll Debug Logging" page; select the "Search Debug" link at the bottom of the page
4 - Select the radio button for the "Detailed Poll Logging by IP" option
4a - Set the IP Address in the IP Address field
4b - If the device resides in a different IP Domain than the default, specify it in the 'IP Domain' field
5 - Select the 'View Data' button to retrieve the data which will display in the web broswer. Leave this browser tab open.
6 - After 20 minutes has passed, refresh the browser tab to display the latest polled data from poll cycles taking place after enabling the logging of poll messages for the device.
7 - Lastly copy and paste the content of the resulting data set (most browsers provide a 'Select All' option off the right click menu on the page) into a simple text doc to attach to the issue for analysis.
8 - Reverse steps 1 and 2 to Disable the debugging using the "Disable Poll Logging" radio button.
If problems are seen with the Ping response, determine the cause and solve the problem.
2 - If there are no problems with Ping between the Data Collector and the device, try running the "Update Metric Family" from the "Polled Metric Families" tab, and/or try the Rediscover button option from the devices Details tab in the Data Aggregator Inventory view.
If none of this helps change the VC to the ICMP VC to provide this data, in the absence of clear Ping response issues between the Data Collector and device, please open a case with CA Support for assistance to determine a root cause and solution. When opening such a case, please reference this Knowledge Base Article (via its ID or full URL) to assist in expediting understanding of the issue at hand.