Add some swap to tox-cover job

As mentioned in the related bug we are seeing
hign memory usage in the job since sqlalchemy
bump to 2.0 and this leads to gate instability.
Adding some swap to handle this until the issue
is fixed.

Also reverts the concurrency reduction change[1]
that was done for this issue as that shouldn't be
needed now.

[1] https://review.opendev.org/c/openstack/neutron/+/920766
Related-Bug: #2065821

Change-Id: I33c4869c1d5230c0c8cc2f0dd4f6d071600bf220

playbooks/configure_swap.yaml

@@ -0,0 +1,3 @@
+- hosts: all
+  roles:
+    - configure-swap

tox.ini

@@ -157,7 +157,7 @@ setenv =
   {[testenv]setenv}
   PYTHON=coverage run --source neutron --parallel-mode
 commands =
-  stestr run --no-subunit-trace --concurrency 4 {posargs}
+  stestr run --no-subunit-trace {posargs}
   coverage combine
   coverage report --fail-under=82 --skip-covered
   coverage html -d cover

zuul.d/job-templates.yaml

@@ -32,6 +32,9 @@
             irrelevant-files:  *irrelevant-files
         - openstack-tox-cover:  # from openstack-cover-jobs template
             timeout: 4800
+            pre-run: playbooks/configure_swap.yaml
+            vars:
+              configure_swap_size: 4096
             irrelevant-files:  *irrelevant-files
         - openstack-tox-py311-with-sqlalchemy-master:
             timeout: 3600