def save_buckets_tree(self, buckets):
ceph_api = CephAPI()
buckets_types = ceph_api.get_bucket_types()
buckets_tree = json.loads(buckets)
crush_map = []
for bucket_tree in buckets_tree:
if "children" in bucket_tree.keys():
get_children_info(bucket_tree, crush_map, buckets_types)
else:
crush_map.append({
"name":
bucket_tree["name"],
"id":
bucket_tree["id"],
"class_ids":
bucket_tree["class_ids"],
"hash":
bucket_tree["hash"],
"alg":
bucket_tree["alg"],
"type_id":
buckets_types.index(bucket_tree["type"])
})
ceph_api.save_buckets(crush_map)
def detach_disk(self, disk_id, pool):
ceph_api = CephAPI()
disk_meta = DiskMeta()
disk_meta.id = disk_id
disk_meta.disk_name = None
return ceph_api.add_disk(disk_meta, False, pool)
def get_disk_snapshots(args):
disk_id = args.disk_id
try:
ceph_api = CephAPI()
image_name = "image-" + disk_id
pool_name = ceph_api.get_pool_bydisk(disk_id)
# If pool inactive #
if pool_name is None:
print("Error : Cannot get pool of disk {}".format(str(disk_id)))
sys.exit(-1)
else:
snaps_ls = ceph_api.get_disk_snapshots(pool_name, image_name)
print(json.dumps(snaps_ls))
sys.exit(0)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : CephException - {} , Cannot get disk {} snapshots.".
format(str(e.message), disk_id))
sys.exit(-1)
except Exception as e:
print("Error : Exception - {}".format(str(e.message)))
sys.exit(-1)
def rollback_snapshot(args):
disk_id = args.disk_id
snap_name = args.snapshot_name
try:
ceph_api = CephAPI()
image_name = str(ceph_api.conf_api.get_image_name_prefix() + disk_id)
pool_name = ceph_api.get_pool_bydisk(
disk_id) # getting pool_name from disk_id
confirm = ceph_api.rollback_to_snapshot(pool_name, image_name,
snap_name)
if not confirm:
print("Error : Cannot rollback disk " + disk_id + " to snapshot " +
snap_name)
sys.exit(-1)
sys.exit(0)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : " + str(e.message) + " , Cannot rollback disk " +
disk_id + " to snapshot " + snap_name)
sys.exit(-1)
except Exception as e:
print("Error : Exception , {}".format(e.message))
sys.exit(-1)
def delete_snapshot(args):
disk_id = args.disk_id
snap_name = args.snapshot_name
try:
ceph_api = CephAPI()
image_name = str(ceph_api.conf_api.get_image_name_prefix() + disk_id)
pool_name = ceph_api.get_pool_bydisk(
disk_id) # getting pool_name from disk_id
# If pool inactive #
if pool_name is None:
print("Error : Cannot get pool of disk {}".format(str(disk_id)))
sys.exit(-1)
else:
confirm = ceph_api.delete_snapshot(pool_name, image_name,
snap_name)
if not confirm:
print("Error : Cannot delete snapshot \"" + snap_name +
"\" of disk " + disk_id)
sys.exit(-1)
sys.exit(0)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : CephException - " + str(e.message) +
" , Cannot delete snapshot \"" + snap_name + "\" of disk " +
disk_id)
sys.exit(-1)
except Exception as e:
print("Error : Exception - {}".format(str(e.message)))
sys.exit(-1)
def get_diskmeta(args):
# Getting a specific disk metadata from "Remote Cluster" giving a disk_id (DiskMeta Objects):
try:
disk_id = args.disk_id
ceph_api = CephAPI()
disk_metadata = ceph_api.get_diskmeta(disk_id)
diskmeta_dict = disk_metadata.__dict__
print(json.dumps(diskmeta_dict, indent=4, sort_keys=False))
sys.exit(0)
except DiskListException as e:
print("Error : DiskListException , {}".format(e.message))
sys.exit(-1)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : {} , Cannot get disk meta.".format(e.message))
sys.exit(-1)
except MetadataException as e:
print("Error : MetadataException , {}".format(e.message))
sys.exit(-1)
except Exception as e:
print("Error : Exception , {}".format(e.message))
sys.exit(-1)
def run(self):
try:
result = Result()
ceph_api = CephAPI()
cluster_status = ceph_api.get_ceph_cluster_status()
if cluster_status is not None:
cluster_status = json.loads(cluster_status)
available_size = 0
used_size = 0
if cluster_status['pgmap']['bytes_total'] > 0:
available_size = cluster_status['pgmap']['bytes_avail'] * 100.0 / cluster_status['pgmap']['bytes_total']
used_size = cluster_status['pgmap']['bytes_used'] * 100.0 / cluster_status['pgmap']['bytes_total']
notify_cluster_space_percent = ConfigAPI().get_notify_cluster_used_space_percent()
if float(used_size) > float(notify_cluster_space_percent):
check_state = self.__context.state.get(self.get_plugin_name(), False)
if check_state == False:
result.title = gettext("core_message_notify_title_cluster_out_space")
result.message = '\n'.join(gettext("core_message_notify_cluster_out_space").split("\\n")).format(int(available_size))
# logger.warning(result.message)
result.plugin_name = str(self.get_plugin_name())
self.__context.results.append(result)
self.__context.state[self.get_plugin_name()] = True
logger.warning("Cluster is running out of disk space")
return
self.__context.state[self.get_plugin_name()] = False
except:
logger.exception("Error occur during get cluster state")
def manage_cluster_setting():
if request.method == 'GET':
result = ""
if list_err in session:
result = session[list_err]
session.pop(list_err)
return render_template('admin/configuration/cluster_settings.html', err=result, form=request.form)
elif list_warning in session:
result = session[list_warning]
session.pop(list_warning)
return render_template('admin/configuration/cluster_settings.html', warning=result, form=request.form)
elif list_success in session:
result = session[list_success]
session.pop(list_success)
cluster_config_model = ClusterConfigForm()
manage_configuration = ManageConfig()
new_cephAPI = CephAPI()
# replica_no = new_cephAPI.get_replicas()
# cluster_config_model.replica_no = replica_no
# compression_algorithm = manage_configuration.get_compression_algorithm()
# cluster_config_model.compression_algorithm = compression_algorithm
# compression_mode = manage_configuration.get_compression_mode()
# cluster_config_model.compression_mode = compression_mode
internet_time_server = manage_configuration.get_ntp_server()
app_config = ManageConfig().get_app_config()
if app_config.email_notify_smtp_server != "":
cluster_config_model.smtp_server = app_config.email_notify_smtp_server
cluster_config_model.smtp_port = app_config.email_notify_smtp_port
cluster_config_model.smtp_email = app_config.email_notify_smtp_email
cluster_config_model.smtp_passwod = app_config.email_notify_smtp_password
cluster_config_model.authentication_value = app_config.email_notify_smtp_security
if internet_time_server == None:
cluster_config_model.internet_time_server = ""
else:
cluster_config_model.internet_time_server = internet_time_server
return render_template('admin/configuration/cluster_settings.html', success=result,
form=cluster_config_model)
else:
cluster_config_model = ClusterConfigForm
manage_configuration = ManageConfig()
new_cephAPI = CephAPI()
# replica_no = new_cephAPI.get_replicas()
# cluster_config_model.replica_no = replica_no
# compression_algorithm = manage_configuration.get_compression_algorithm()
# cluster_config_model.compression_algorithm = compression_algorithm
# compression_mode = manage_configuration.get_compression_mode()
# cluster_config_model.compression_mode = compression_mode
internet_time_server = manage_configuration.get_ntp_server()
app_config = ManageConfig().get_app_config()
if app_config.email_notify_smtp_server != "":
cluster_config_model.smtp_server = app_config.email_notify_smtp_server
cluster_config_model.smtp_port = app_config.email_notify_smtp_port
cluster_config_model.smtp_email = app_config.email_notify_smtp_email
cluster_config_model.smtp_passwod = app_config.email_notify_smtp_password
cluster_config_model.authentication_value = app_config.email_notify_smtp_security
if internet_time_server == None:
cluster_config_model.internet_time_server = ""
else:
cluster_config_model.internet_time_server = internet_time_server
return render_template('/admin/configuration/cluster_settings.html', form=cluster_config_model)
def delete_snapshots(args):
disk_id = args.disk_id
try:
ceph_api = CephAPI()
image_name = str(ceph_api.conf_api.get_image_name_prefix() + disk_id)
pool_name = ceph_api.get_pool_bydisk(
disk_id) # getting pool_name from disk_id
confirm = ceph_api.delete_snapshots(pool_name, image_name)
if not confirm:
print("Error : Cannot delete all snapshots of disk " + disk_id)
sys.exit(-1)
sys.exit(0)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : " + str(e.message) +
" , Cannot delete all snapshots of disk " + disk_id)
sys.exit(-1)
except Exception as e:
print("Error : Exception , {}".format(e.message))
sys.exit(-1)
def get_disks_meta(self):
ceph_api = CephAPI()
consul_api = ConsulAPI()
ls = ceph_api.get_disks_meta()
for disk in ls:
if disk and hasattr(disk, "paths") and not disk.paths:
disk.status = DisplayDiskStatus.unattached
elif disk and hasattr(disk, "paths") and disk.paths:
data = consul_api.find_disk(disk.id)
if data is not None:
disk.status = DisplayDiskStatus.starting
if str(data.Flags) == "1":
disk.status = DisplayDiskStatus.stopping
elif consul_api.is_path_locked(disk.id):
disk.status = DisplayDiskStatus.started
else:
disk.status = DisplayDiskStatus.stopped
job_manager = JobManager()
job_list = job_manager.get_running_job_list()
for j in job_list:
# Check if the status running
if j.is_running:
# Set disk status [deleting]
if j.type == JobType.DELETE_DISK and str(
j.params).find(str(disk.id)) > -1:
disk.status = DisplayDiskStatus.deleting
return ls
def set_compression_mode(self, mode, pool="rbd"):
api = CephAPI()
if mode == CompressionMode.none:
mode_str = 'none'
elif mode == CompressionMode.force:
mode_str = 'force'
return api.set_compression_mode(mode_str, pool)
def start(self, disk_id, pool):
try:
ceph_api = CephAPI()
consul_api = ConsulAPI()
attr = ceph_api.read_image_metadata(
ConfigAPI().get_image_name_prefix() + disk_id, pool)
petasan_meta = attr.get(ConfigAPI().get_image_meta_key())
disk_meta = DiskMeta()
if petasan_meta:
disk_meta.load_json(petasan_meta)
else:
return Status.error
consul_api.add_disk_resource(disk_meta.id, "disk")
consul_api.add_disk_pool(disk_meta.id, pool)
i = 0
for p in disk_meta.paths:
i += 1
consul_api.add_disk_resource(
"/".join(["", disk_meta.id, str(i)]), None)
except Exception as e:
logger.error("Can not start disk %s" % disk_id)
logger.exception(e.message)
return Status.error
return Status.done
def get_templates(self):
templates = {}
path = ConfigAPI().get_crush_rule_templates_path()
ceph_api = CephAPI()
next_id = ceph_api.get_next_rule_id()
for f in os.listdir(path):
if os.path.isdir(path + f):
continue
with open(path + f, 'r') as file:
lines = file.readlines()
body = ''
for line in lines:
if len(line) == 0:
continue
if line.startswith('id '):
body += '# Auto generated id. do not modify \n'
body += 'id ' + next_id + '\n'
continue
body += line
templates[f] = body
return templates
def get_compression_mode(self, pool='rbd'):
api = CephAPI()
mode = api.get_compression_mode(pool)
if mode is None or mode is '' or "none" in mode:
return CompressionMode.none
if 'force' in mode:
return CompressionMode.force
raise Exception("Compression mode is not valid")
def get_list():
ceph_api = CephAPI()
for i in ceph_api.get_disks_meta():
print i.user, i.disk_name, i.password
for p in i.get_paths():
print p.ip, p.subnet_mask, p.eth
def set_disk_metadata(args):
io_ctx = None
ceph_api = CephAPI()
cluster = None
try:
cluster = ceph_api.connect()
io_ctx = cluster.open_ioctx(args.pool)
# Get which ceph user is using this function & get his keyring file path #
ceph_auth = CephAuthenticator()
config = configuration()
cluster_name = config.get_cluster_name()
if args.file:
with open(str(args.file), 'r') as file:
disk_metadata_str = file.read()
else:
disk_metadata = sys.stdin.readlines()
disk_metadata_str = ''.join(
str(line)
for line in disk_metadata) # converting list to string
# read object meta :
cmd = "rbd info " + args.pool + "/" + str(
args.image) + " " + ceph_auth.get_authentication_string(
) + " --cluster " + cluster_name + " | grep rbd_data"
ret, stdout, stderr = exec_command_ex(cmd)
if ret != 0:
if stderr:
cluster.shutdown()
print("Cannot get image meta object from rbd header.")
rbd_data = stdout.rstrip().strip()
dot_indx = rbd_data.rfind(".")
image_id = rbd_data[(dot_indx + 1):]
meta_object = "rbd_header." + image_id
attr_object = meta_object
io_ctx.set_xattr(str(attr_object),
str(ConfigAPI().get_image_meta_key()),
disk_metadata_str)
io_ctx.close()
cluster.shutdown()
sys.exit(0)
except Exception as e:
print("Error in executing script function : set_disk_metadata , " +
str(e.message))
io_ctx.close()
cluster.shutdown()
sys.exit(-1)
def get_active_replicated_pools(self):
ceph_api = CephAPI()
active_pools = ceph_api.get_active_pools()
pools_info = ceph_api.get_pools_info()
replicated_active_pools = []
for pool in pools_info:
if pool.type == 'replicated' and pool.name in active_pools:
replicated_active_pools.append(pool.name)
return replicated_active_pools
def rados_benchmark(self, block_size, is_type_write, duration_sec, threads,
pool):
ceph_api = CephAPI()
if is_type_write:
return ceph_api.rados_write(duration_sec, threads, block_size,
pool)
else:
return ceph_api.rados_read(duration_sec, threads, pool)
def read_disks_metadata(args):
io_ctx = None
ceph_api = CephAPI()
cluster = None
try:
cluster = ceph_api.connect()
io_ctx = cluster.open_ioctx(args.pool)
# Get which ceph user is using this function & get his keyring file path #
ceph_auth = CephAuthenticator()
config = configuration()
cluster_name = config.get_cluster_name()
cmd = "rbd info " + args.pool + "/" + str(
args.image) + " " + ceph_auth.get_authentication_string(
) + " --cluster " + cluster_name + " | grep rbd_data"
ret, stdout, stderr = exec_command_ex(cmd)
if ret != 0:
if stderr:
cluster.shutdown()
print("Cannot get image meta object from rbd header.")
sys.exit(-1)
rbd_data = stdout.rstrip().strip()
dot_indx = rbd_data.rfind(".")
image_id = rbd_data[(dot_indx + 1):]
rbd_header_object = "rbd_header." + image_id
try:
ret = io_ctx.get_xattr(rbd_header_object, meta_key)
except:
ret = io_ctx.get_xattr(rbd_header_object[:-1], meta_key)
io_ctx.close()
cluster.shutdown()
if ret:
print(ret)
sys.stdout.flush()
sys.exit(0)
else:
# Non-PetaSAN Disk :
sys.exit(-1)
except Exception as e:
print("Error in executing script function : read_disks_metadata , " +
str(e.message))
io_ctx.close()
cluster.shutdown()
sys.exit(-1)
def __clean_unused_rbd_images(self):
ceph_api = CephAPI()
rbd_images = ceph_api.get_mapped_images()
if rbd_images is None:
return
for image, mapped_count in rbd_images.iteritems():
if image not in self.__backstore:
if int(mapped_count) > 0:
for i in range(0, int(mapped_count)):
ceph_api.unmap_image(image)
logger.debug("Unmapped unused image {}.".format(image))
def get_meta():
ceph_confg = CephAPI()
ceph_api = CephAPI()
logger.info("done , create disk")
attr = ceph_api.read_image_metadata("000001")
xx = attr.get(app_conf.get_image_meta_key())
logger.info(xx)
disk = DiskMeta()
disk.load_json(xx)
logger.info("disk user is %s" % (disk.user))
def get_pool_by_disk(disk_id):
consul_api = ConsulAPI()
ceph_api = CephAPI()
pool = consul_api.get_disk_pool(disk_id)
if pool:
return pool
pool = ceph_api.get_pool_bydisk(disk_id)
if pool:
return pool
return None
def set_compression_algorithm(self, algorithm, pool="rbd"):
api = CephAPI()
if algorithm == CompressionAlgorithm.none:
algorithm_str = 'none'
elif algorithm == CompressionAlgorithm.zlib:
algorithm_str = 'zlib'
elif algorithm == CompressionAlgorithm.snappy:
algorithm_str = 'snappy'
elif algorithm == CompressionAlgorithm.zstd:
algorithm_str = 'zstd'
elif algorithm == CompressionAlgorithm.lz4:
algorithm_str = 'lz4'
return api.set_compression_algorithm(algorithm_str, pool)
def stop_all_disks():
ceph_api = CephAPI()
disks = ceph_api.get_disks_meta()
if disks is None:
print('Error reading disk metadata from Ceph')
return False
disks.sort(key=lambda disk: disk.id)
for disk_meta in disks:
if is_disk_started(disk_meta.id):
if stop_disk(disk_meta.id, disk_meta.pool):
print "Stopped disk " + disk_meta.id + " " + disk_meta.disk_name
else:
print "Error stopping disk " + disk_meta.id + " " + disk_meta.disk_name
return True
def get_compression_algorithm(self, pool="rbd"):
api = CephAPI()
algorithm = api.get_compression_algorithm(pool)
if algorithm is None or algorithm is '' or "none" in algorithm:
return CompressionAlgorithm.none
if 'zlib' in algorithm:
return CompressionAlgorithm.zlib
if 'snappy' in algorithm:
return CompressionAlgorithm.snappy
if 'zstd' in algorithm:
return CompressionAlgorithm.zstd
if 'lz4' in algorithm:
return CompressionAlgorithm.lz4
raise Exception("Compression algorithm is not valid")
def run(self):
try:
result = Result()
ceph_status_overall = ""
ceph_api = CephAPI()
cluster_status = ceph_api.get_ceph_cluster_status() # ceph status --format json-pretty
if cluster_status is not None:
cluster_status = json.loads(cluster_status)
# Ceph 12 :
if "overall_status" in cluster_status["health"] and cluster_status["health"]["overall_status"] is not None:
ceph_status_overall = cluster_status["health"]["overall_status"]
else:
ceph_status_overall = cluster_status["health"]["status"]
if ceph_status_overall == "HEALTH_ERR":
prv_err = self.__context.state.get(self.get_plugin_name(), False)
if not prv_err:
ceph_health_obj = cluster_status["health"]
summary_messages = ""
summary_messages_ls = []
if "checks" in ceph_health_obj:
for key in ceph_health_obj["checks"]:
if ceph_health_obj["checks"][key] is not None:
msg = ceph_health_obj["checks"][key]["summary"]["message"]
summary_messages_ls.append(msg)
summary_messages = '\n '.join(summary_messages_ls)
result.title = gettext("core_message_notify_cluster_status_title")
result.message = '\n'.join(gettext("core_message_notify_cluster_status_body").split("\\n")).format(summary_messages)
result.plugin_name = str(self.get_plugin_name())
self.__context.results.append(result)
self.__context.state[self.get_plugin_name()] = True
logger.warning("Cluster overall health status is HEALTH_ERR")
return
self.__context.state[self.get_plugin_name()] = False
except Exception as e:
logger.exception(e)
logger.error("An error occurred while ClusterStatusPlugin was running.")
def validate_new_iscsi2_ip(self, ip):
if not self.is_ip_in_iscsi2_subnet(ip):
return NewIPValidation.wrong_subnet
ips_used = set()
# get the image metada from ceph
ceph_api = CephAPI()
for i in ceph_api.get_disks_meta():
for p in i.get_paths():
ips_used.add(p.ip)
if ip in ips_used:
return NewIPValidation.used_already
return NewIPValidation.valid
def get_disk_paths(self, disk_id, pool):
paths_list = CephAPI().get_disk_meta(disk_id, pool).paths
paths_list_with_node = []
sessions_dict = ConsulAPI().get_sessions_dict(
ConfigAPI().get_iscsi_service_session_name())
# in case consul lock on disk
for kv in ConsulAPI().get_disk_paths(disk_id):
path = Path()
path_str = paths_list[int(str(kv.Key).split(disk_id + "/")[1]) - 1]
path.load_json(json.dumps(path_str))
if hasattr(kv, "Session") and sessions_dict.has_key(kv.Session):
path.locked_by = sessions_dict.get(kv.Session).Node
paths_list_with_node.append(path)
else:
paths_list_with_node.append(path)
# in case disk is stopped
if not paths_list_with_node:
for path_str in paths_list:
path = Path()
path.load_json(json.dumps(path_str))
paths_list_with_node.append(path)
return paths_list_with_node
def get_all_images(args):
try:
ceph_api = CephAPI()
pool_name = args.pool
images_ls = ceph_api.get_all_images(pool_name)
print(json.dumps(images_ls))
sys.exit(0)
except CephException as e:
if e.id == CephException.GENERAL_EXCEPTION:
print("Error : {} , Cannot get images list.".format(e.message))
sys.exit(-1)
except Exception as e:
print("Error : Exception , {}".format(e.message))
sys.exit(-1)
def _get_pool_by_disk(self, disk_id):
consul_api = ConsulAPI()
ceph_api = CephAPI()
pool = consul_api.get_disk_pool(disk_id)
if pool:
logger.info('Found pool:{} for disk:{} via consul'.format(
pool, disk_id))
return pool
pool = ceph_api.get_pool_bydisk(disk_id)
if pool:
logger.info('Found pool:{} for disk:{} via ceph'.format(
pool, disk_id))
return pool
logger.error('Could not find pool for disk ' + disk_id)
return None