Uploaded image for project: 'Lustre'
  1. Lustre
  2. LU-19

imperative recovery

XMLWordPrintable

    • Icon: New Feature New Feature
    • Resolution: Fixed
    • Icon: Blocker Blocker
    • Lustre 2.2.0
    • Lustre 2.0.0, Lustre 1.8.6
    • None
    • 18,767
    • 10457

      Imperative recovery means the clients are notified explicitly when and where a failed target has
      restarted. Ideally the notification should occur after the target has mounted and is waiting for
      connections. A fully automated version of this will require a new mechanism for distributing node
      health information efficiently. A simple, initial version can be done as a /proc file for each
      import that can be used to initiate recovery.

        1. Imperative Recovery Design Document.pdf
          428 kB
        2. Imperative Recovery Scope Statement.pdf
          109 kB
        3. Imperative Recovery Test Plan Rev A.pdf
          217 kB
        4. LUG11_Imperative_Recovery.pdf
          399 kB
        1.
        		 find connection uuid by peer nid Technical task Resolved Jinshan Xiong (Inactive)
        2.
        		 Remove Parallel AST limits Technical task Resolved Jinshan Xiong (Inactive)
        3.
        		 Make ptlrpc_recover_import async Technical task Resolved Jinshan Xiong (Inactive)
        4.
        		 conf-sanity test_41b: @@@@@@ FAIL: test_41b failed with 1 Technical task Resolved Jinshan Xiong (Inactive)
        5.
        		 Use cfs_rwlock_t for obd_dev_lock instead of spinlock Technical task Resolved WC Triage
        6.
        		 export ldlm_lock_set_data and implement it correctly Technical task Resolved Jinshan Xiong (Inactive)
        7.
        		 update mgc llog process code Technical task Resolved WC Triage

            jay Jinshan Xiong (Inactive)
            jay Jinshan Xiong (Inactive)
            Votes:
            0 Vote for this issue
            Watchers:
            7 Start watching this issue

              Created:
              Updated:
              Resolved: