Fixed databricks_job update from instance pool to regular node type

[marekbrysa]

Changes

When a job cluster was updated from instance pool to regular, the computed driver_instance_pool_id was still being sent, which then caused failure of the cluster to start with error in the UI "Worker must use a pool if driver uses a pool".

This resets the driver_instance_pool_id if instance_pool_id is not set.

Also I noticed there was a function FixInstancePoolChangeIfAny that was only used on databricks_cluster. Looking at #850, I believe it's also relevant for jobs clusters, so applying it too.

Tests

Tested updating a jobs cluster from regular to pool and back.

• make test run locally
• relevant change in docs/ folder
• covered with integration tests in internal/acceptance
• relevant acceptance tests are passing
• using Go SDK

[codecov-commenter]

Codecov Report

Merging #2549 (af1f2b0) into master (2873d96) will increase coverage by 0.00%.
Report is 3 commits behind head on master.
The diff coverage is 83.33%.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #2549   +/-   ##
=======================================
  Coverage   88.18%   88.19%           
=======================================
  Files         145      145           
  Lines       12013    12050   +37     
=======================================
+ Hits        10594    10627   +33     
- Misses        933      935    +2     
- Partials      486      488    +2     

Files Changed	Coverage Δ
clusters/clusters_api.go	85.30% <71.42%> (-0.41%)	⬇️
clusters/resource_cluster.go	85.58% <100.00%> (+0.48%)	⬆️
jobs/resource_job.go	91.66% <100.00%> (+0.08%)	⬆️

... and 1 file with indirect coverage changes

[mgyucht]

Nice! Any chance you can do a manual test of this to verify?

[marekbrysa]

@mgyucht yeah, I already did, it seems to work: "Tested updating a jobs cluster from regular to pool and back."