ceph-ansible-pull-requests: add os_tuning_params to --extra-vars

Set these to the same values used for vagrant testing. The specific
issue here is that the defaults for os_tuning_params includes setting
vm.min_free_kbytes to 4194303 which is more memory than the jenkins
slave used for testing has.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>

diff --git a/ceph-ansible-pull-requests/build/build b/ceph-ansible-pull-requests/build/build
index 3fdd76740e96d4d49c7cf80d373fabd5c84a74a6..4f9fd3ab8283967ae0c27186ced7abfa529889bf 100644 (file)
--- a/ceph-ansible-pull-requests/build/build
+++ b/ceph-ansible-pull-requests/build/build
@@ -9,4 +9,4 @@ install_python_packages "pkgs[@]"
 source $VENV/activate
 
 cd "$WORKSPACE"/ceph-ansible
-$VENV/ansible-playbook -vv -i tests/inventories/single-machine.yml -c local test.yml --extra-vars='{"devices":["/dev/vdb"], "ceph_stable":true, "journal_collocation":true, "journal_size":1024, "monitor_interface":"eth1", "cluster_network":"127.0.0.1/0", "public_network":"127.0.0.1/0"}'
+$VENV/ansible-playbook -vv -i tests/inventories/single-machine.yml -c local test.yml --extra-vars='{"devices":["/dev/vdb"], "ceph_stable":true, "journal_collocation":true, "journal_size":1024, "monitor_interface":"eth1", "cluster_network":"127.0.0.1/0", "public_network":"127.0.0.1/0"}, "os_tuning_params":[{"name": "kernel.pid_max", "value": 4194303}, {"name": "fs.file-max", "value": 26234859}]'