#27734: Parallel test workers can be assigned an unexpected index when 
recovering
from errors
-----------------------------------+------------------------------------
     Reporter:  Adam Wróbel        |                    Owner:  nobody
         Type:  Bug                |                   Status:  new
    Component:  Testing framework  |                  Version:  1.10
     Severity:  Normal             |               Resolution:
     Keywords:  parallel           |             Triage Stage:  Accepted
    Has patch:  0                  |      Needs documentation:  0
  Needs tests:  0                  |  Patch needs improvement:  0
Easy pickings:  0                  |                    UI/UX:  0
-----------------------------------+------------------------------------
Changes (by Jacob Walls):

 * has_patch:  1 => 0
 * summary:
     Add a helpful error message when a parallel test worker is assigned an
     unexpected index
     =>
     Parallel test workers can be assigned an unexpected index when
     recovering from errors
 * type:  Cleanup/optimization => Bug

Comment:

 Thanks Aarav Sharma for alerting me that we had an issue for this. I've
 been meaning to open this exact issue for a while.

 It's easiest to simulate the ''occasional'' failure described here by just
 forcing workers to respawn with `maxtasksperchild`:

 {{{#!diff
 diff --git a/django/test/runner.py b/django/test/runner.py
 index 56eda9d98d..ddbbeee510 100644
 --- a/django/test/runner.py
 +++ b/django/test/runner.py
 @@ -576,6 +576,9 @@ class ParallelTestSuite(unittest.TestSuite):
                  self.debug_mode,
                  self.used_aliases,
              ],
 +            # For testing only,
 +            # Emulate the occasional failures that happen (under resource
 contention?)
 +            maxtasksperchild=1,
          ) as pool:
              test_results =
 pool.imap_unordered(self.run_subsuite.__func__, args)

 }}}

 Then run a subsuite of tests that make heavy use of the database:

 `./runtests.py queries`

 {{{#!py
   File "/Users/jwalls/django/django/db/backends/sqlite3/creation.py", line
 146, in setup_worker_connection
     source_db = self.connection.Database.connect(
         f"file:{source_db_name}?mode=ro", uri=True
     )
 sqlite3.OperationalError: unable to open database file
 }}}

 ----
 I'd like to reframe this ticket to actually fix the underlying issue
 instead of adjusting the error message. We could do that by making the
 database clones a pool that that workers pull from instead of trying to
 map ephemeral workers 1:1 with fixed resources.  All of this functionality
 in the parallel runner is private, so we shouldn't be shy about fixing the
 design.
-- 
Ticket URL: <https://code.djangoproject.com/ticket/27734#comment:15>
Django <https://code.djangoproject.com/>
The Web framework for perfectionists with deadlines.

-- 
You received this message because you are subscribed to the Google Groups 
"Django updates" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion visit 
https://groups.google.com/d/msgid/django-updates/0107019d21c15aa7-aca50fc8-be54-4658-8367-a7d3575bb206-000000%40eu-central-1.amazonses.com.

Reply via email to