feat(AWS Deploy): Support disableRollback parameter

[fredericbarthelet]

Closes: #10235

[codecov]

Codecov Report

Merging #10236 (f2f8f7a) into master (44511f3) will decrease coverage by 0.02%.
The diff coverage is 40.00%.

❗ Current head f2f8f7a differs from pull request most recent head d75cc99. Consider uploading reports for the commit d75cc99 to get more accurate results

@@            Coverage Diff             @@
##           master   #10236      +/-   ##
==========================================
- Coverage   85.39%   85.37%   -0.03%     
==========================================
  Files         339      339              
  Lines       13842    13843       +1     
==========================================
- Hits        11821    11818       -3     
- Misses       2021     2025       +4     

Impacted Files	Coverage Δ
lib/plugins/aws/provider.js	94.31% <ø> (ø)
lib/plugins/aws/lib/updateStack.js	94.80% <40.00%> (-3.81%)	⬇️
lib/cli/triage.js	90.80% <0.00%> (-1.34%)	⬇️
lib/plugins/aws/package/compile/events/sqs.js	100.00% <0.00%> (ø)

---

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 44511f3...d75cc99. Read the comment docs.

[pgrzesik]

Thanks @fredericbarthelet - looks good, I only have minor remarks, let me know what do you think 🙇

[pgrzesik]

Well done, thank you @fredericbarthelet 🙇

[RichardBradley]

Why did you write "for non-production environment" here?

I think this param is to control a choice as to whether to attempt to fix a partial deployment, or to roll it all back. That's not necessarily tied to production or not.

"true" is required to work around AWS issues like aws-cloudformation/cloudformation-coverage-roadmap#573 , whether in production or non-production

(see also https://github.com/serverless/serverless/blame/f22354e8dd45871915e21e5656055a9213f93ff0/docs/providers/aws/guide/serverless.yml.md#L95 )

Terraform has no-rollback as its default behaviour and plenty of people use that in production