LU-18273 tests: do reboot and failback in several attempts

LU-18273 tests: do reboot and failback in several attempts

Instead of playing around the ha_failback_delay values
different for different clusters and to avoid the tests
failures unrelated to Lustre - it looks reasonable just to
try "failover/failback" several times.

Test-Parameters: trivial
Signed-off-by: Elena Gryaznova <elena.gryaznova@hpe.com>
HPE-bug-id: LUS-11816
Reviewed-by: Vladimir Saveliev <vladimir.saveliev@hpe.com>
Reviewed-by: Alexander Boyko <alexander.boyko@hpe.com>
Change-Id: I748e4e580408f7662b2c64576af36637cfe46ef3
Reviewed-on: https://review.whamcloud.com/c/fs/lustre-release/+/56514
Tested-by: jenkins <devops@whamcloud.com>
Tested-by: Maloo <maloo@whamcloud.com>
Reviewed-by: Oleg Drokin <green@whamcloud.com>