Build Docker once to share with all E2E jobs

[awrichar]

Each E2E job in the workflow matrix currently takes ~10 min, where the first ~5 min is just spent building the FireFly Docker image.

By building the Docker once upfront and sharing it across all jobs, each test job can be cut in half. This shaves 20-30 min off the combined runtime of the workflow jobs and makes more efficient use of the available workers.

As a small side benefit, the Docker image for each run is also stored as an artifact available for download.

[codecov-commenter]

Codecov Report

Merging #909 (203e550) into main (8e17c9a) will decrease coverage by 0.00%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main     #909      +/-   ##
==========================================
- Coverage   99.97%   99.97%   -0.01%     
==========================================
  Files         298      297       -1     
  Lines       19473    19398      -75     
==========================================
- Hits        19469    19393      -76     
  Misses          4        4              
- Partials        0        1       +1     

Impacted Files	Coverage Δ
internal/tokens/fftokens/fftokens.go	99.56% <0.00%> (-0.44%)	⬇️
pkg/core/event.go	100.00% <0.00%> (ø)
pkg/core/namespace.go	100.00% <0.00%> (ø)
internal/data/data_manager.go	100.00% <0.00%> (ø)
internal/namespace/manager.go	100.00% <0.00%> (ø)
internal/multiparty/manager.go	100.00% <0.00%> (ø)
internal/definitions/handler.go	100.00% <0.00%> (ø)
internal/orchestrator/status.go	100.00% <0.00%> (ø)
internal/reference/reference.go	100.00% <0.00%> (ø)
internal/events/network_action.go	100.00% <0.00%> (ø)
... and 9 more

Help us with your feedback. Take ten seconds to tell us how you rate us.

[shorsher]

🐳

[shorsher]

should we also pull in #873?