Uploaded image for project: 'WiredTiger'
  1. WiredTiger
  2. WT-964

Make placement of LSM chunks configurable

    • Type: Icon: Task Task
    • Resolution: Duplicate
    • Priority: Icon: Major - P3 Major - P3
    • None
    • Affects Version/s: None
    • Component/s: None

      At the moment, LSM just uses an incrementing chunk ID to generate the URI for each chunk and Bloom filter. If we made an interface for this, something like:

      int generate_uri(..., int chunk_id, int level, int type, char *buf, size_t *len)
      

      where type == WT_LSM_DATA || type == WT_LSM_BLOOM

      Then applications could place Bloom filters and chunks in different places, or place different levels of the tree on different types of storage subsystem.

            Assignee:
            backlog-server-execution [DO NOT USE] Backlog - Storage Execution Team
            Reporter:
            michael.cahill@mongodb.com Michael Cahill (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

              Created:
              Updated:
              Resolved: