Restructure Analysis Metadata Fetching for Celery

[ilan-gold]

Speed up Celery by storing results of multiple API calls (that don't change) in a table / using existing stored information instead of making the request. This addresses both #2642 and #3435. The speed up is drastic, and gets larger (in absolute terms) the more requests that are made. Benchmarking is in #3435 but it appears that the speed up is on the magnitude of 5x or greater for large calls.

[codecov]

Codecov Report

Merging #3461 into develop will decrease coverage by 0.81%.
The diff coverage is 84.61%.

@@             Coverage Diff             @@
##           develop    #3461      +/-   ##
===========================================
- Coverage    88.22%   87.41%   -0.82%     
===========================================
  Files           96       96              
  Lines        17170    14796    -2374     
===========================================
- Hits         15149    12934    -2215     
+ Misses        2021     1862     -159

Impacted Files	Coverage Δ
refinery/tool_manager/tests.py	99.89% <100%> (-0.04%)	⬇️
refinery/analysis_manager/tasks.py	72.04% <100%> (-0.1%)	⬇️
refinery/tool_manager/models.py	95.32% <78.57%> (-0.64%)	⬇️
refinery/data_set_manager/views.py	65.71% <0%> (-4.65%)	⬇️
refinery/core/views.py	69.24% <0%> (-3.4%)	⬇️
refinery/core/models.py	82.47% <0%> (-3.39%)	⬇️
refinery/tool_manager/views.py	92.3% <0%> (-0.82%)	⬇️
refinery/analysis_manager/test_tasks.py	100% <0%> (ø)	⬆️
... and 3 more

---

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 66f92d5...63106a0. Read the comment docs.