GPT Data Generator: explore new ways to tune connection pool size depending on environment size
With Gitaly Cluster coming soon, we need to explore new approaches to tune connection pool size depending on environment size to achieve a good balance between speed and stability in data generation with threads. Currently it's being done by taking Storage node count into account as larger environments have more Gitaly nodes. One of the solutions could be to specify target environment type similarly to target_user_count
option in #335.
Follow-up on #380 (closed)
Edited by Nailia Iskhakova