fix: scale number of threads to scale with number of cpus

[igorbernstein2]

This is temporary solution until something better can be worked out.
Ideally we would separate the executor for gax and the transport provider. But there are some backwards compatibility concerns to workout first.

[codecov]

Codecov Report

Merging #878 into master will increase coverage by 0.02%.
The diff coverage is 100%.

@@             Coverage Diff              @@
##             master     #878      +/-   ##
============================================
+ Coverage      78.6%   78.63%   +0.02%     
- Complexity     1162     1163       +1     
============================================
  Files           203      203              
  Lines          5141     5143       +2     
  Branches        413      413              
============================================
+ Hits           4041     4044       +3     
  Misses          925      925              
+ Partials        175      174       -1

Impacted Files	Coverage Δ	Complexity Δ
...le/api/gax/core/InstantiatingExecutorProvider.java	62.5% <100%> (+5.35%)	3 <1> (ø)	⬇️
.../java/com/google/api/gax/batching/BatcherImpl.java	98% <0%> (+0.66%)	16% <0%> (+1%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 327cd8d...9864ee8. Read the comment docs.

[vam-google]

LGTM

(with fingers crossed)

[igorbernstein2]

It seems the presubmits are broken and are preventing merging prs

[igorbernstein2]

@vam-google would you mind cutting a release with this fix?

[vam-google]

I believe DPE team now handles this. @chingor13 can you cut a release?

[codyoss]

@igorbernstein2 Does Runtime.getRuntime().availableProcessors() work as expected in kubernetes container environments? I personally don't know, but a quick search yielded some people having issues where it returns 1 for them.

[igorbernstein2]

Not really. If there is no cpu config, then it will default to 1, which is why the PR sets a minimum to 4 as before. This is just a bandaid, the real fix would be more along the lines of #869. However that approach was a bit controversial, so this change was put in place until @vam-google could figure out how to incorporate #869 without too much breakage