def are_all_self_heal_daemons_are_online(mnode, volname):
"""Verifies whether all the self-heal-daemons are online for the specified
volume.
Args:
mnode (str): Node on which cmd has to be executed.
volname (str): volume name
Returns:
bool : True if all the self-heal-daemons are online for the volume.
False otherwise.
NoneType: None if unable to get the volume status
"""
from glustolibs.gluster.volume_libs import is_distribute_volume
if is_distribute_volume(mnode, volname):
g.log.info(
"Volume %s is a distribute volume. "
"Hence not checking for self-heal daemons "
"to be online", volname)
return True
service = 'shd'
failure_msg = ("Verifying all self-heal-daemons are online failed for "
"volume %s" % volname)
# Get volume status
vol_status = get_volume_status(mnode=mnode,
volname=volname,
service=service)
if vol_status is None:
g.log.error(failure_msg)
return None
# Get all nodes from pool list
from glustolibs.gluster.peer_ops import nodes_from_pool_list
all_nodes = nodes_from_pool_list(mnode)
if not all_nodes:
g.log.error(failure_msg)
return False
online_status = True
for node in all_nodes:
node_shd_status_value = (
vol_status[volname][node]['Self-heal Daemon']['status'])
if node_shd_status_value != '1':
online_status = False
g.run(mnode, ("gluster volume status %s shd" % volname))
if online_status is True:
g.log.info("All self-heal Daemons are online")
return True
else:
g.log.error("Some of the self-heal Daemons are offline")
return False
def wait_for_self_heal_daemons_to_be_online(mnode, volname, timeout=300):
"""Waits for the volume self-heal-daemons to be online until timeout
Args:
mnode (str): Node on which commands will be executed.
volname (str): Name of the volume.
Kwargs:
timeout (int): timeout value in seconds to wait for self-heal-daemons
to be online.
Returns:
True if all self-heal-daemons are online within timeout,
False otherwise
"""
from glustolibs.gluster.volume_libs import is_distribute_volume
if is_distribute_volume(mnode, volname):
g.log.info(
"Volume %s is a distribute volume. "
"Hence not waiting for self-heal daemons "
"to be online", volname)
return True
counter = 0
flag = 0
while counter < timeout:
status = are_all_self_heal_daemons_are_online(mnode, volname)
if status:
flag = 1
break
if not status:
time.sleep(10)
counter = counter + 10
if not flag:
g.log.error(
"All self-heal-daemons of the volume '%s' are not online "
"even after %d minutes", (volname, timeout / 60.0))
return False
else:
g.log.info("All self-heal-daemons of the volume '%s' are online ",
volname)
return True