Optimize the memory usage of certain graphs

[sighingnow]

What do these changes do?

• Expose the retain_oid options to Python
• Upgrade vineyard to latest version to handle the memory usage issues
• Adding errors="ignore" to avoid decoding/encoding errors
• Enable int32_t OID type support

Benchmarking

Settings:

• 1 worker
• loading from CSV (710MB)
  • nodes: 330MB, 21865475
  • edges: 380MB, 11089373
• workload: loading (generate_eid=False) -> project -> pagerank

Memory usage:

• int64_t oid:
  • retain_oid=True:
    • after loading graph, peak: 3.53GB
    • after project, peak: 3.72GB
    • after running the application, peak: 4.49GB
  • retain_oid=False:
    • after loading graph, peak: 3.33GB
    • after project, peak: 3.53GB
    • after running the application, peak: 4.29GB
• string oid: graph: , peak:
  • retain_oid=True:
    • after loading graph, peak: 5.16GB
    • after project, peak: 5.18GB
    • after running the application, peak: 6.00GB
  • retain_oid=False:
    • after loading graph, peak: 4.68GB
    • after project, peak: 4.70GB
    • after running the application, peak: 5.51GB

Related issue number

Fixes #2127
Fixes #2269

See also: v6d-io/v6d#1151

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov-commenter]

Codecov Report

Merging #2381 (953ccbf) into main (e5aed5a) will increase coverage by 38.05%.
The diff coverage is 89.02%.

Additional details and impacted files

@@             Coverage Diff             @@
##             main    #2381       +/-   ##
===========================================
+ Coverage   34.38%   72.43%   +38.05%     
===========================================
  Files          88       88               
  Lines        9711     9752       +41     
===========================================
+ Hits         3339     7064     +3725     
+ Misses       6372     2688     -3684     

Impacted Files	Coverage Δ
python/graphscope/analytical/app/java_app.py	24.36% <ø> (-4.57%)	⬇️
python/graphscope/client/connection.py	38.61% <0.00%> (ø)
python/graphscope/framework/dag_utils.py	65.89% <ø> (+30.79%)	⬆️
python/graphscope/client/session.py	75.37% <40.00%> (+13.34%)	⬆️
python/graphscope/framework/graph_builder.py	88.57% <50.00%> (ø)
python/graphscope/tests/unittest/test_lazy.py	96.21% <50.00%> (+96.21%)	⬆️
python/graphscope/framework/graph.py	85.37% <80.00%> (+25.69%)	⬆️
python/graphscope/tests/conftest.py	82.71% <92.00%> (+34.83%)	⬆️
python/graphscope/client/archive.py	64.15% <100.00%> (+28.30%)	⬆️
python/graphscope/config.py	100.00% <100.00%> (ø)
... and 70 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 97b7224...953ccbf. Read the comment docs.

[yecol]

Responsive patch! 💐

[github-actions]

😭 Deploy PR Preview 45cec05 failed. Build logs

_{🤖 By surge-preview}