Server uses sync.Pool for gzip.Writer allocations

[michel-laterman]

Use of gzip.Writer pool for plain HTTP responses - gzip.Writers have a large memory footprint, if many agent clients clients were to respond at once a server could have an OOM crash. Use of the pool also reduces allocs.

go test -v -benchmem -benchtime 5s -bench=BenchmarkCompressGzip
...
goos: darwin
goarch: arm64
pkg: github.com/open-telemetry/opamp-go/server
cpu: Apple M3 Pro
BenchmarkCompressGzip
BenchmarkCompressGzip/with_pool
BenchmarkCompressGzip/with_pool-12         	  483432	     12155 ns/op	     113 B/op	       2 allocs/op
BenchmarkCompressGzip/no_pool
BenchmarkCompressGzip/no_pool-12           	   96456	     60972 ns/op	  813988 B/op	      19 allocs/op

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.77%. Comparing base (d508718) to head (2594b4b).
⚠️ Report is 10 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #484      +/-   ##
==========================================
+ Coverage   81.71%   81.77%   +0.06%     
==========================================
  Files          27       27              
  Lines        2122     2129       +7     
==========================================
+ Hits         1734     1741       +7     
  Misses        263      263              
  Partials      125      125              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[tigrannajaryan]

Please break this down into individual PRs, one PR per topic.

use of gzip.Writer pool for plain HTTP responses - gzip.Writers have a large memory footprint, if many agent clients clients were to respond at once a server could have an OOM crash. Use of the pool also reduces allocs.

Please add a benchmark that shows before/after memory/CPU usage and demonstrates the improvement.

[michel-laterman]

Callback fix pr: #488

[michel-laterman]

non gwPool changes: #489

I'm going to rewrite this PR's history so that the 1st commit has both compress methods, and a benchmark for them, and the 2nd commit only has the (new) pool method.