model: make ID fields non-pointer string types

[axw]

Motivation/summary

Make more ID fields non-pointer types.

I missed a subtle issue with #3849, which changed some code to store a pointer to an empty parent ID in some cases. This didn't cause any change in the fields output, because we check != nil and != "" when producing beat events. However, when aggregating transaction metrics, we are only checking != nil.

To simplify the logic, and avoid these kinds of errors in the future, just store a string value. As a side benefit, this reduces heap allocations.

$ benchstat /tmp/old.txt /tmp/new.txt
name                                          old time/op    new time/op    delta                                                                                                                          
BackendProcessor/heavy.ndjson/NoRateLimit-12    38.4ms ± 4%    33.2ms ± 7%  -13.51%  (p=0.008 n=5+5)                                                                                                       
                                                                                                                                                                                                           
name                                          old speed      new speed      delta                                                                                                                          
BackendProcessor/heavy.ndjson/NoRateLimit-12  10.4MB/s ± 4%  12.0MB/s ± 7%  +15.69%  (p=0.008 n=5+5)                                                                                                       
                                                                                                                                                                                                           
name                                          old alloc/op   new alloc/op   delta                                                                                                                          
BackendProcessor/heavy.ndjson/NoRateLimit-12    6.34MB ± 0%    6.33MB ± 0%     ~     (p=0.151 n=5+5)
                                                                                                     
name                                          old allocs/op  new allocs/op  delta
BackendProcessor/heavy.ndjson/NoRateLimit-12      113k ± 0%      112k ± 0%   -1.27%  (p=0.008 n=5+5)

Included is a tangentially related change to stop recording the parent span index, used for RUMv3, on the model types. We now just use the index during decoding and then discard it.

Checklist

• I have signed the Contributor License Agreement.
  - [ ] I have updated CHANGELOG.asciidoc

I have considered changes for:
- [ ] documentation
- [ ] logging (add log lines, choose appropriate log selector, etc.)
- [ ] metrics and monitoring (create issue for Kibana team to add metrics to visualizations, e.g. Kibana#44001)

• automated tests (add tests for the code changes, all unit tests pass locally)
  - [ ] telemetry
  - [ ] Elasticsearch Service (https://cloud.elastic.co)
  - [ ] Elastic Cloud Enterprise (https://www.elastic.co/products/ece)
  - [ ] Elastic Cloud on Kubernetes (https://www.elastic.co/elastic-cloud-kubernetes)

How to test these changes

make test

Related issues

None.

[apmmachine]

💚 Build Succeeded

Expand to view the summary

Build stats

• Build Cause: [Pull request #3853 updated]

• Start Time: 2020-06-08T06:42:36.850+0000

• Duration: 46 min 28 sec

Test stats 🧪

Test	Results
Failed	0
Passed	3204
Skipped	146
Total	3350