[Varnish] #1733: Unstable test-cases (was: test l00002 is not stable)

[Varnish]

#1733: Unstable test-cases
-------------------------+---------------------
 Reporter:  phk          |       Owner:  martin
     Type:  defect       |      Status:  new
 Priority:  normal       |   Milestone:
Component:  varnishtest  |     Version:  trunk
 Severity:  normal       |  Resolution:
 Keywords:               |
-------------------------+---------------------
Description changed by phk:

Old description:

> the pipelined requests gets varying XID's making the logexpects fail

New description:

 After spending some time on r01086's stability I ran an overnight '-n
 1000' test run on both project.v-c.o and my own lab machine.

 The set of tests failing on the two machines are almost distinct:

 {{{
 p.v-c.o: Linux 3.2.0-83-virtual x86_64
   11 tests/r01086.vtc
    4 tests/r00942.vtc
    2 tests/c00067.vtc
    2 tests/c00039.vtc
    2 tests/b00046.vtc
    1 tests/r01441.vtc
    1 tests/r00962.vtc
    1 tests/r00861.vtc
    1 tests/l00002.vtc

 ni.f.d: FreeBSD 11.0-CURRENT amd64
  205 tests/l00002.vtc
   32 tests/s00004.vtc
   23 tests/c00046.vtc
   12 tests/c00066.vtc
    7 tests/l00005.vtc
    2 tests/c00049.vtc
    2 tests/c00023.vtc
    1 tests/r01494.vtc
    1 tests/r01030.vtc
    1 tests/c00045.vtc
    1 tests/c00020.vtc
    1 tests/c00004.vtc
    1 tests/c00002.vtc
    1 tests/c00001.vtc
    1 tests/b00005.vtc
    1 tests/b00002.vtc
 }}}

 Considering that the total run involved nearly half a million tests on
 each machine, the overall failure-rate is quite acceptable, but it is
 still obvious that we have test-cases which are not entirely stable.

--

-- 
Ticket URL: <https://www.varnish-cache.org/trac/ticket/1733#comment:1>
Varnish <https://varnish-cache.org/>
The Varnish HTTP Accelerator