def _recover_network_isolation(self, cluster_name, fail_node_name):
cluster = ResourceManager.get_cluster(cluster_name)
if not cluster:
logging.error("RecoverManager : cluster not found")
return
fail_node = cluster.get_node_by_name(fail_node_name)
logging.info(
"RecoverManager : start to recover the network isolation of node %s"
% (fail_node))
self._recover_vm(cluster, fail_node)
return self._recover_node_by_reboot(fail_node)
def _recover_hardware_fault(self, cluster_name, fail_node_name):
cluster = ResourceManager.get_cluster(cluster_name)
if not cluster:
logging.error("RecoverManager : cluster not found")
return
fail_node = cluster.get_node_by_name(fail_node_name)
logging.info(
"RecoverManager : start to recover the hardware fault of node %s" %
(fail_node))
self._recover_vm(cluster, fail_node)
return self._recover_node_by_reboot(fail_node)
def _recover_power_off(self, cluster_name, fail_node_name):
cluster = ResourceManager.get_cluster(cluster_name)
if not cluster:
logging.error("RecoverManager : cluster not found")
return
fail_node = cluster.get_node_by_name(fail_node_name)
print "fail node is %s, power fail" % fail_node.name
print "start recovery vm"
self._recover_vm(cluster, fail_node)
print "end recovery vm"
return self._recover_node_by_start(fail_node)
def _recover_os_hanged(self, cluster_name, fail_node_name):
cluster = ResourceManager.get_cluster(cluster_name)
if not cluster:
logging.error("RecoverManager : cluster not found")
return
fail_node = cluster.get_node_by_name(fail_node_name)
print "fail node is %s, OS fail" % fail_node.name
logging.info("fail node is %s, OS fail", fail_node.name)
print "start recovery vm"
self._recover_vm(cluster, fail_node)
print "end recovery vm"
return self._recover_node_by_reboot(fail_node)
def _recover_instance_crash(self, cluster_name, failed_instance_name):
cluster = ResourceManager.get_cluster(cluster_name)
if not cluster:
logging.warning(
"RecoverManager, _recover_instance_crash: cluster not found")
return
failed_instance = cluster.get_protected_instance_by_instance_name(
failed_instance_name)
if failed_instance == None:
logging.warning(
"RecoverManager, _recover_instance_crash: failed instance not found"
)
return
logging.info(
"RecoverManager : start to recover the instance crash of instance %s"
% (failed_instance_name))
result = failed_instance.recover_instance_crash()
return result
