Hi, I’m evaluating harness on a project we have here and am running into an issue with the stackdriver integration for pulling in log data.
The stackdriver integration tests fine, and is able to pull data for both the hostname and message fields using the “Guide from example”, but is unable to do anything with “Expression for host/container name” and the documentation also fails to cover what this is meant to do.
What format is this meant to be taking and what exactly is it supposed to be referring to? The test button is also greyed out and submitting with the default and then running a deployment fails in verification with no request logged and a STACKDRIVER_ERROR response.
The Expression for Host/Container Name is how Harness knows which host to look at for the analysis based on the way that your Stackdriver Metrics show the name of the Host or Container.
If you do a query of the metrics for one of the recent deployments, check the metadata for that deployment in Stackdriver and see what the key:value pair is which tells you the name of the host or container.
If you do a query of the metrics for one of the recent deployments
So this is actually for log verification, not the metrics so I can’t do that. I’m looking at the log data and trying relevant fields but the feedback is unfortunately not helping me figure out the problem or what is expected(where is the data for the guide coming from? Is this assuming stackdriver is collecting logs from a gke cluster?).
By running a deployment with no verification, the “guide from example” is now working and I was able to try a few values(which all exist in the example guilde), but running the test would consistently not work with “Verification Data Not Found” and “No Data found for the selected node”
The 3rd party API call in the test is succeeding with multiple log entries in the response. So I presume that the host/container name is supposed to be filtering relevant entries only from these but it seems to be filtering everything?
left is the guide from example for the hostname
middle top for the host/container name
middle bottom is the settings I actually ran the test on
and the right is a section of the api response.
The host selected for the test configuration was “filter-platform-gateway-78c86fb495-kxlt2”
If you can help clarify what’s going on here, it’d be very appreciated.
Quick question, but do you have any traffic hitting that/those particular pod(s) during the 15 minutes after deployment? If not, then there will be no data found.
Also, if you go to the executed workflow and select the verification box, you should be able to click on the hamburger menu on the right side and select 3rd party calls. What do you see there?
