Uploaded image for project: 'WiredTiger'
  1. WiredTiger
  2. WT-13370

Reuse kv_database/kv_table inside the test/model workload generator

    • Type: Icon: Task Task
    • Resolution: Unresolved
    • Priority: Icon: Major - P3 Major - P3
    • None
    • Affects Version/s: None
    • Component/s: Test Model
    • Storage Engines
    • StorEng - Defined Pipeline

      test/model's workload generator currently uses an ad-hoc mechanism for keeping track of which keys exist in a table, so that it can choose existing keys such as for the evict and the remove operators. The problem is that this mechanism is quite imperfect, and can't account for changes in key existence during events such as transaction rollbacks or database crashes, which prevents it from being able to create certain kinds of interesting workloads.

      The workload generator should thus use kv_database and kv_table for this purpose. This would also require us to implement a mechanism to choose an existing random within kv_table.

            Assignee:
            backlog-server-storage-engines [DO NOT USE] Backlog - Storage Engines Team
            Reporter:
            peter.macko@mongodb.com Peter Macko
            Votes:
            0 Vote for this issue
            Watchers:
            2 Start watching this issue

              Created:
              Updated: