Thundra tries to affect function size at a minimum level since our Node.JS library only has an additional size of 6.5 MB.
You can use manual instrumentation to create spans for code that you would like to monitor. Instructions for performing manual instrumentation for the Node.js agent can be found here.
You can add application and invocation tags using our tagging. On the Query Bar you can input the tags you added to filter your functions/invocations.
For example, let's say that you have functions with different versions and you need to query a specific version. You can add the application tag version as the key, and value as a string value, such as "1.2.1". The environment variable should be as follow:
thundra_agent_lambda_application_tag_version:"1.2.1"
Then, on the query bar on Function List page you can search with that filter:
If you want to search through invocations using a label, you can add invocation tags (see an example here). For our example above, the tag we set should be as follows:
thundra.InvocationSupport.setTag('version', '1.2.1');
Next, search on the Query Bar on the Function Details page using that filter.
You can use our sampling feature (described here). By configuring intelligent sampling, you can sample your data without losing interesting data. You can also have a look at this blog to find more information about intelligent sampling.
​