3534709538
At the moment there is a race condition possible on repo container startup, when glusterd.service is started later then var-www-repo.mount attempts to. This leads to a mount failure and requires a manual intervention after each controller restart. We use BindsTo together with After to ensure that mount will be started only when glusterd.service reaches Active state. A related fix [1] was proposed a while ago, stating that glusterd.service is started before network-online.target, which is not a case, as glusterd also starts after network-online.target [1] https://review.opendev.org/c/openstack/openstack-ansible-plugins/+/845987 Related-Bug: #2104382 Change-Id: I19c95e48eafbecbb8a6bc8bc7b2bf2e7d6b09333 Signed-off-by: Dmitriy Rabotyagov <dmitriy.rabotyagov@advanced.host>