VAULT-31907: Entity loading speedup

[miagilepner]

Description

This PR has 3 performance improvements in the loadEntities function:

1. Batch memDB operations into groups of 1024. Previously, we were creating a new transaction per entity.
2. Cache the bucket reads from the local alias storage packer and use those when loading entities
3. Deserialize the entities concurrently

I ran benchmarks using:

• 200k total entities
• Each entity had 1 local alias and 1 non-local alias

I got the number 1024 by batching entities together across buckets (which this branch does not implement, but I used for benchmarking so that I wouldn't need to create millions of entities). These batches are counted by adding the number of entities and the number of aliases per entity (note: this is different than what I posted on slack earlier today. I was incorrectly only counting the number of entities per batch).

Benchmark_loadEntities/tx_size_2048-10         	       5	22934412175 ns/op
Benchmark_loadEntities/tx_size_1024-10         	       5	6821009883 ns/op
Benchmark_loadEntities/tx_size_512-10          	       5	6938667308 ns/op
Benchmark_loadEntities/tx_size_256-10          	       5	6962929416 ns/op
Benchmark_loadEntities/tx_size_1-10            	       5	10874873183 ns/op

I also benchmarked the other 2 performance improvements; cache_aliases_{true|false} being optimization number 2, and unmarshal_parallel_{true|false} being optimization number 3.

The results of the benchmarks are:

Benchmark_loadEntities/tx_size_1024_cache_aliases_true_unmarshal_parallel_true_-10         	       3	5570892945 ns/op
Benchmark_loadEntities/tx_size_1024_cache_aliases_true_unmarshal_parallel_false_-10        	       3	6047840347 ns/op
Benchmark_loadEntities/tx_size_1024_cache_aliases_false_unmarshal_parallel_true_-10        	       3	5858072153 ns/op
Benchmark_loadEntities/tx_size_1024_cache_aliases_false_unmarshal_parallel_false_-10       	       3	6072485403 ns/op
Benchmark_loadEntities/tx_size_1_cache_aliases_true_unmarshal_parallel_true_-10            	       3	9174082680 ns/op
Benchmark_loadEntities/tx_size_1_cache_aliases_true_unmarshal_parallel_false_-10           	       3	9550873292 ns/op
Benchmark_loadEntities/tx_size_1_cache_aliases_false_unmarshal_parallel_true_-10           	       3	9403697055 ns/op
Benchmark_loadEntities/tx_size_1_cache_aliases_false_unmarshal_parallel_false_-10          	       3	9821321805 ns/op

TODO only if you're a HashiCorp employee

• Backport Labels: If this fix needs to be backported, use the appropriate backport/ label that matches the desired release branch. Note that in the CE repo, the latest release branch will look like backport/x.x.x, but older release branches will be backport/ent/x.x.x+ent.
  • LTS: If this fixes a critical security vulnerability or severity 1 bug, it will also need to be backported to the current LTS versions of Vault. To ensure this, use all available enterprise labels.
• ENT Breakage: If this PR either 1) removes a public function OR 2) changes the signature
  of a public function, even if that change is in a CE file, double check that
  applying the patch for this PR to the ENT repo and running tests doesn't
  break any tests. Sometimes ENT only tests rely on public functions in CE
  files.
• Jira: If this change has an associated Jira, it's referenced either
  in the PR description, commit message, or branch name.
• RFC: If this change has an associated RFC, please link it in the description.
• ENT PR: If this change has an associated ENT PR, please link it in the
  description. Also, make sure the changelog is in this PR, not in your ENT PR.

[github-actions]

CI Results:
All Go tests succeeded! ✅

[github-actions]

Build Results:
All builds succeeded! ✅

[digivava]

Great clarity in the changelog entry

[kubawi]

Agreed!

[kubawi]

LGTM!