Uploaded image for project: 'vpp'
  1. vpp
  2. VPP-1785

PAPI is not reliable enough for scale tests

XMLWordPrintable

    • Icon: Bug Bug
    • Resolution: Done
    • Icon: Medium Medium
    • 20.01
    • None
    • Python API binding
    • None

      During the CSIT work [0] of using PAPI (over socket) for scale tests, there are failures. The symptoms are consistent with either VPP not sending a response (or not receiving the command), or sending responses out of order. The failure happens fairly infrequently per command (say 1 on 10000), but scale tests execute high amount of commands.

      It is not clear what is happening on VPP side. Perhaps the cause is the same as for VPP-1753.

      The cause is unlikely to be happening in binary APIs, as in CSIT the currently run scale test are using VAT which does not fail.

      [0] https://gerrit.fd.io/r/c/csit/+/21876

            vrpolak Vratko Polak
            vrpolak Vratko Polak
            Votes:
            0 Vote for this issue
            Watchers:
            1 Start watching this issue

              Created:
              Updated:
              Resolved: