Uploaded image for project: 'Core Server'
  1. Core Server
  2. SERVER-40856

Log replication progress in hang analyzer

    XMLWordPrintableJSON

Details

    • Server Tooling & Methods

    Description

      Some repl hangs are due to nodes not replicating rather than an actual deadlock. It would be helpful if the hang analyzer called replSetGetStatus on every node in the cluster while the process was still alive.

      If the replSetGetStatus call hangs because of a deadlock on the ReplicationCoordinator mutex, then the replication progress is probably not important anyways so it's not a problem to just kill that command.

      Attachments

        Activity

          People

            backlog-server-stm Backlog - Server Tooling and Methods (STM) (Inactive)
            judah.schvimer@mongodb.com Judah Schvimer
            Votes:
            0 Vote for this issue
            Watchers:
            3 Start watching this issue

            Dates

              Created:
              Updated: