Uploaded image for project: 'Indy Node'
  1. Indy Node
  2. INDY-1152

Failed restart after getting unhandled exception (KeyError)

    XMLWordPrintable

Details

    • Bug
    • Status: Complete
    • Medium
    • Resolution: Done
    • None
    • None
    • None
    • None
    • indy-node 1.2.50-stable

    • Sprint 18.03 Stability, DKMS, Sprint 18.04

    Description

      While testing pool upgrade was noted, that nodes which failed the upgrade procedure had broken indy-node processes. In systemd service we have option like 'Restart=on-failure' that meens that service must be restarted when running process will broke (by KeyError unhandled exception for example). All of nodes with faulty upgrade procedure had unhandled exception and systemctl did not restart indy-node service. 
      The main assumption is that main python process had unhandled exception, was stopped but child processes was not stopped and systemd not restart service.
      Ways to investigate:

      1. There is several type of service - 'forked' and 'sample'. What of type do we use dy default?
      2. How many subprocesses we use for indy-node service

      Also, set 'SIGTERM' into list of handled signal.

      Attachments

        1. INDY-1152.PNG
          328 kB
          Vladimir Shishkin

        Issue Links

          Activity

            People

              VladimirWork Vladimir Shishkin
              anikitinDSR Andrew Nikitin
              Andrew Nikitin, Vladimir Shishkin
              Votes:
              0 Vote for this issue
              Watchers:
              2 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: