def get_vms(session_id, page_size):
vm_filter_spec = VirtualMachineFilterSpec()
vm_filter_spec.live = "true"
vm_filter_spec.limit = page_size
# Repeat with the live filter.
live_vms = tintri.get_vms(filters=vm_filter_spec)
if live_vms.filteredTotal == 0:
raise TintriServerError(0, cause="No live VMs present")
print_info("Live Total = " + str(live_vms.filteredTotal))
print ""
count = 1
for vm in live_vms:
vm_name = vm.vmware.name
vm_uuid = vm.uuid.uuid
if debug_mode:
dump_object(vm.stat.sortedStats[0])
vm_stats = VmStat(vm_name, vm_uuid, vm.stat.sortedStats[0])
print_debug(str(count) + ": " + vm_name + ", " + vm_uuid)
count += 1
# Store the VM stats object keyed by VM name.
vms[vm_name] = vm_stats
return vms
def main():
"""
Main function
"""
args = get_args()
output_bucket = args.pathoutput
storage = args.storage
full_table_path = args.bqtable
# Fetch training data from BQ or GCS
if storage in ['BQ', 'bq' 'bigquery', 'BigQuery', 'bigQuery', 'Bigquery', 'Bq']:
dataset = utils.read_df_from_bigquery(full_table_path)
else:
dataset = utils.get_data_from_gcs(args.pathdata)
x_train, y_train, x_val, y_val = utils.data_train_test_split(dataset)
# Get pipeline and fit model
pipeline = model.get_pipeline()
pipeline.fit(x_train, y_train)
scores = model_selection.cross_val_score(pipeline, x_val, y_val, cv=3)
print('model score: %.3f' % pipeline.score(x_val, y_val))
print('pipeline run done :)')
# Output results and trained model
model_output_path = os.path.join(output_bucket,'model/', metadata.MODEL_FILE_NAME)
metric_output_path = os.path.join(output_bucket, 'experiment', metadata.METRIC_FILE_NAME)
utils.dump_object(pipeline, model_output_path)
utils.dump_object(scores, metric_output_path)
joblib_output_path = os.path.join(output_bucket)
# Write GCS path to local file
# This GCS path can be passed to the next component
Path(args.pathoutputfile).parent.mkdir(parents=True, exist_ok=True)
Path(args.pathoutputfile).write_text(joblib_output_path)
def main():
args = get_args()
path_data = args.pathdata
output_bucket = args.pathoutput
storage = args.storage
full_table_path = args.bqtable
if storage in [
'BQ', 'bq'
'bigquery', 'BigQuery', 'bigQuery', 'Bigquery', 'Bq'
]:
dataset = utils.read_df_from_bigquery(full_table_path)
else:
dataset = utils.get_data_from_gcs(path_data)
x_train, y_train, x_val, y_val = utils.data_train_test_split(dataset)
pipeline = model.get_pipeline()
pipeline.fit(x_train, y_train)
scores = model_selection.cross_val_score(pipeline, x_val, y_val, cv=3)
print("model score: %.3f" % pipeline.score(x_val, y_val))
print('pipeline run done :)')
model_output_path = os.path.join(output_bucket, 'model/',
metadata.MODEL_FILE_NAME)
metric_output_path = os.path.join(output_bucket, 'experiment',
metadata.METRIC_FILE_NAME)
utils.dump_object(pipeline, model_output_path)
utils.dump_object(scores, metric_output_path)
joblib_output_path = os.path.join(output_bucket)
Path(args.pathoutputfile).parent.mkdir(parents=True, exist_ok=True)
Path(args.pathoutputfile).write_text(joblib_output_path)
def clone_vm(uuid, clone_name, vcenter_name, datastore_name, num_clones):
# Create and initialize the clone spec minus the UUID
clone_spec = VirtualMachineCloneSpec()
clone_spec.consistency = 'CRASH_CONSISTENT'
clone_spec.count = num_clones
vmware_clone_info = VMwareCloneInfo()
clone_spec.vmware = vmware_clone_info
clone_spec.vmware.cloneVmName = clone_name
clone_spec.vmware.vCenterName = vcenter_name
clone_spec.vmware.datastoreName = datastore_name
# Suss-out if UUID is a Tintri VM or snapshot UUID and set the appropriate field.
if (uuid.find("VIM") > -1):
clone_spec.vmId = uuid
elif (uuid.find("SST") > -1):
clone_spec.snapshotId = uuid
else:
raise TintriServerError(0,
cause="UUID is not a VM or snapshot UUID: " +
uuid)
print_info("Cloning " + uuid + " to " + clone_name + " for " +
str(num_clones) + " times")
if debug_mode:
dump_object(clone_spec)
# Clone the VM specified by the clone_payload
task_result = tintri.clone_vm(clone_spec, True)
# Get the desired task info
task_uuid = task_result.uuid.uuid
task_state = task_result.state
task_progress = task_result.progressDescription
task_type = task_result.type
print_info(task_type + ": " + task_uuid + ": " + task_state + " - " +
task_progress)
def dump(self, filename):
dump_object(filename, self.pack())
from config_tools import load_config, build_graph, create_index_from_config
from utils import dump_object, print_object
from index import diff_trees
from graph import depth_first_search, find_roots
import json
# index1 = create_index_from_config(".apps/alcuin1.json")
index1 = create_index_from_config(".apps/MDC_ENSSUP_EVO.json")
# index2 = create_index_from_config(".apps/alcuin1_other.json")
index2 = create_index_from_config(".apps/MODULE_ENSSUP.json")
(id1, t1) = index1.write_tree()
(id2, t2) = index2.write_tree()
t = dict(t1, **t2)
dump_object("trees.json", t)
dump_object('i1.json', dict(index1.blobs, **index2.blobs))
def visitor(*args):
print(args)
diff_trees(visitor, id1, id2, t)
def _check_params(self):
app_log.debug("jdbe:check:start")
utils.dump_object(self)
# raise TypeError, "Parameter error"
app_log.debug("jdbe:check:complete")