Description of problem: Our official documentation recommends to increase restart delays for large Ceph clusters [1] (merged as a solution for bug #1620699). Basically, we recommend the customer to set the following parameters using THT: parameter_defaults: CephAnsibleExtraConfig: health_osd_check_delay: 40 health_osd_check_retries: 30 health_mon_check_retries: 10 health_mon_check_delay: 20 The truth is that this configuration change is not a silver bullet and doesn't actually work for the bug #1620699 itself: specified parameters are hard-coded in rolling_update.yml (it is reasonable high there) and switch-from-non-containerized-to-containerized-ceph-daemons.yml (quite low there) playbooks. I understand that this issue should be likely handled by ceph-ansible (we can increase hard-coded values) or documentation (we can tell customer to adjust playbook), but wanted to ask THT developers to make a first touch and decide which way will work for us here. [1] https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/13/html/fast_forward_upgrades/assembly-preparing_for_overcloud_upgrade#increasing-the-restart-delay-for-large-ceph-clusters
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2020:0760