def setup_class(self):
"""Setup class parameters"""
# create the tile index table. skip if it exists
try:
TileIndexDB.createTable(endpoint_url=settings.DYNAMO_ENDPOINT)
CuboidIndexDB.createTable(endpoint_url=settings.DYNAMO_ENDPOINT)
except Exception as e:
pass
self.tileindex_db = TileIndexDB(nd_proj.project_name,
endpoint_url=settings.DYNAMO_ENDPOINT)
self.tile_bucket = TileBucket(nd_proj.project_name,
endpoint_url=settings.S3_ENDPOINT)
[self.x_tile, self.y_tile, self.z_tile] = [0, 0, 0]
supercuboid_key = 'testing'
message_id = '123456'
receipt_handle = 'testing123456'
message = serializer.encodeDeleteMessage(supercuboid_key, message_id,
receipt_handle)
# insert message in the upload queue
CleanupQueue.createQueue(nd_proj, endpoint_url=settings.SQS_ENDPOINT)
self.cleanup_queue = CleanupQueue(nd_proj,
endpoint_url=settings.SQS_ENDPOINT)
self.cleanup_queue.sendMessage(message)
# receive message and upload object
for z_index in range(self.z_tile, settings.SUPER_CUBOID_SIZE[2], 1):
tile_handle = cStringIO.StringIO()
self.tile_bucket.putObject(tile_handle, nd_proj.channel_name,
nd_proj.resolution, self.x_tile,
self.y_tile, z_index, message_id,
receipt_handle)
def teardown_class(self):
"""Teardown class parameters"""
# cleanup tilebucket
for z_index in (self.z_tile, settings.SUPER_CUBOID_SIZE[2], 1):
tile_key = self.tile_bucket.encodeObjectKey(
nd_proj.channel_name,
nd_proj.resolution,
self.x_tile,
self.y_tile,
z_index,
)
self.tile_bucket.deleteObject(tile_key)
morton_index = XYZMorton(self.tiles)
supercuboid_key = self.cuboid_bucket.generateSupercuboidKey(
nd_proj.channel_name, nd_proj.resolution, self.tiles)
self.cuboid_bucket.deleteObject(supercuboid_key)
# delete created entities
TileIndexDB.deleteTable(endpoint_url="http://localhost:8000")
CuboidIndexDB.deleteTable(endpoint_url="http://localhost:8000")
IngestQueue.deleteQueue(nd_proj, endpoint_url="http://localhost:4568")
CleanupQueue.deleteQueue(nd_proj, endpoint_url="http://localhost:4568")
TileBucket.deleteBucket(endpoint_url="http://localhost:4567")
try:
CuboidBucket.deleteBucket(endpoint_url="http://localhost:4567")
except Exception as e:
pass
def setup_class(self):
"""Setup parameters"""
try:
CuboidIndexDB.createTable(endpoint_url=settings.DYNAMO_ENDPOINT)
except Exception as e:
pass
self.cuboid_index = CuboidIndexDB(
nd_proj.project_name, endpoint_url=settings.DYNAMO_ENDPOINT)
def __init__(self, pr):
"""Create the database connection"""
self.pr = pr
# create connections for cuboid bucket and cuboid dyanmo table
self.cuboid_bucket = CuboidBucket(
self.pr.project_name, endpoint_url=ndingest_settings.S3_ENDPOINT)
self.cuboidindex_db = CuboidIndexDB(
self.pr.project_name,
endpoint_url=ndingest_settings.DYNAMO_ENDPOINT)
def __init__(self, token_name, host_name=HOST_NAME):
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}_upload.log'.format(token_name))
self.logger.addHandler(fh)
self.info_interface = InfoInterface(host_name, token_name)
self.project_name = self.info_interface.project_name
self.cuboidindex_db = CuboidIndexDB(self.project_name)
self.cuboid_bucket = CuboidBucket(self.project_name)
def setup_class(self):
"""Setup class parameters"""
# create the tile index table. skip if it exists
try:
TileIndexDB.createTable(endpoint_url="http://localhost:8000")
CuboidIndexDB.createTable(endpoint_url="http://localhost:8000")
except Exception as e:
pass
self.tileindex_db = TileIndexDB(nd_proj.project_name,
endpoint_url="http://localhost:8000")
# create the tile bucket
TileBucket.createBucket(endpoint_url="http://localhost:4567")
self.tile_bucket = TileBucket(nd_proj.project_name,
endpoint_url="http://localhost:4567")
self.tiles = [self.x_tile, self.y_tile, self.z_tile] = [0, 0, 0]
message_id = "testing"
receipt_handle = "123456"
# insert SUPER_CUBOID_SIZE tiles in the bucket
for z_index in (self.z_tile, settings.SUPER_CUBOID_SIZE[2], 1):
tile_handle = cStringIO.StringIO()
self.tile_bucket.putObject(
tile_handle,
nd_proj.channel_name,
nd_proj.resolution,
self.x_tile,
self.y_tile,
z_index,
message_id,
receipt_handle,
)
# creating the cuboid bucket
CuboidBucket.createBucket(endpoint_url="http://localhost:4567")
self.cuboid_bucket = CuboidBucket(nd_proj.project_name,
endpoint_url="http://localhost:4567")
# create the ingest queue
IngestQueue.createQueue(nd_proj, endpoint_url="http://localhost:4568")
self.ingest_queue = IngestQueue(nd_proj,
endpoint_url="http://localhost:4568")
# send message to the ingest queue
morton_index = XYZMorton(self.tiles)
supercuboid_key = self.cuboid_bucket.generateSupercuboidKey(
nd_proj.channel_name, nd_proj.resolution, morton_index)
response = self.ingest_queue.sendMessage(supercuboid_key)
# create the cleanup queue
CleanupQueue.createQueue(nd_proj, endpoint_url="http://localhost:4568")
def __init__(self, token_name, host_name=HOST_NAME):
"""Create the bucket and intialize values"""
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}.log'.format(token_name))
self.logger.addHandler(fh)
# setting up the project metadata
self.info_interface = InfoInterface(host_name, token_name)
# creating the resource interface to the remote server
# self.resource_interface = ResourceInterface(self.info_interface.dataset_name, self.info_interface.project_name, host_name, logger=self.logger)
# self.proj = self.resource_interface.getProject()
# create the s3 I/O and index objects
self.cuboidindex_db = CuboidIndexDB(self.info_interface.project_name)
self.cuboid_bucket = CuboidBucket(self.info_interface.project_name)
def __init__(self, db):
"""Connect to the S3 backend"""
try:
self.db = db
self.project_name = self.db.proj.project_name
self.cuboidindex_db = CuboidIndexDB(self.project_name)
self.cuboid_bucket = CuboidBucket(self.project_name)
except Exception, e:
logger.error("Cannot connect to S3 backend")
raise SpatialDBError("Cannot connect to S3 backend")
def createTestDB ( project_name, channel_list=['unit_anno'], channel_type=ANNOTATION, channel_datatype=UINT32, public=PUBLIC_TRUE, ximagesize=10000, yimagesize=10000, zimagesize=1000, xvoxelres=4.0, yvoxelres=4.0, zvoxelres=3.0, scalingoption=ZSLICES, scalinglevels=5, readonly=READONLY_FALSE, propagate=NOT_PROPAGATED, window=[0,0], time=[0,15], default=False, nd_version=ND_VERSION, token_name='unittest', user='******', dataset_name="unittest", base_resolution=0):
"""Create a unit test data base on the specified sit and name"""
# setting s3backend to true if Redis and creating s3 bucket and dynamo table
if KV_ENGINE == REDIS:
s3backend = S3_TRUE
CuboidIndexDB.createTable()
CuboidBucket.createBucket()
else:
s3backend = S3_FALSE
unituser = User.objects.get(username=user)
ds = NDDataset(Dataset ( dataset_name=dataset_name, user=unituser, ximagesize=ximagesize, yimagesize=yimagesize, zimagesize=zimagesize, xoffset=0, yoffset=0, zoffset=1, xvoxelres=xvoxelres, yvoxelres=yvoxelres, zvoxelres=zvoxelres, scalingoption=scalingoption, scalinglevels=scalinglevels, public=PUBLIC_TRUE, dataset_description="Unit test" ) )
ds.create()
# make the project entry
pr = NDProject(Project(project_name=project_name, project_description='Unit test', user=unituser, dataset=ds._ds, nd_version=nd_version, host='localhost', kvengine=KV_ENGINE, kvserver=KV_SERVER, s3backend=s3backend))
pr.create()
# create a token
tk = NDToken(Token (token_name = token_name, user = unituser, token_description = 'Unit test token', project_id = pr.project_name, public = public))
tk.create()
# get the correct object for the kvengine
# pd = NDProjectsDB.getProjDB(NDProjectpr)
# create the database
# pd.newNDProject()
try:
for channel_name in channel_list:
ch = NDChannel(Channel (channel_name=channel_name, channel_type=channel_type, channel_datatype=channel_datatype, channel_description='Unit test channel', project_id=pr.project_name, readonly=readonly, propagate=propagate, resolution=base_resolution, exceptions=1, starttime=time[0], endtime=time[1] ,startwindow=window[0], endwindow=window[1], default=default))
# create a channel
ch.create()
# create the channel table
# pd.newNDChannel(ch.channel_name)
except Exception, e:
print(e)
raise e
def deleteTestDB ( project_name, token_name='unittest' ):
try:
# get the objects
tk = NDToken.fromName(token_name)
tk.delete()
pr = NDProject.fromName(project_name)
ds = pr.datasetcfg
# tk = Token.objects.get(token_name=token_name)
# pr = Project.objects.get(project_name=project_name)
# ds = Dataset.objects.get(dataset_name=pr.dataset_id)
# get the channel list
# channel_list = Channel.objects.filter(project_id=pr)
# get the correct object for the kvengine
# pd = NDProjectsDB.getProjDB(pr)
for ch in pr.projectChannels():
ch.delete()
# delete the channel table
# pd.deleteNDChannel(ch.channel_name)
# delete the channel
# ch.delete()
# delete the project database
# pd.deleteNDProject()
# delete the objects
pr.delete()
ds.delete()
# delete s3 bucket and dynamo table
if KV_ENGINE == REDIS:
CuboidIndexDB.deleteTable()
CuboidBucket.deleteBucket()
except Exception, e:
print(e)
raise e
class S3Cuboid(object):
def __init__(self, token_name, host_name=HOST_NAME):
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}_upload.log'.format(token_name))
self.logger.addHandler(fh)
self.info_interface = InfoInterface(host_name, token_name)
self.project_name = self.info_interface.project_name
self.cuboidindex_db = CuboidIndexDB(self.project_name)
self.cuboid_bucket = CuboidBucket(self.project_name)
def upload(self, file_name, channel_name, resolution, x_index, y_index, z_index, dimensions=[1, 64, 512,512], time_index=0, neariso=False):
"""Upload a 4D supercuboid directly to dynamo and s3"""
cuboid_data = np.fromfile(file_name, dtype=self.info_interface.get_channel_datatype(channel_name))
cuboid_data = cuboid_data.reshape(dimensions)
super_zidx = XYZMorton([x_index, y_index, z_index])
self.logger.info("Inserting cube {},{},{}".format(x_index, y_index, z_index))
self.cuboidindex_db.putItem(channel_name, resolution, x_index, y_index, z_index, time_index, neariso=neariso)
self.cuboid_bucket.putObject(channel_name, resolution, super_zidx, time_index, blosc.pack_array(cuboid_data), neariso=neariso)
def deleteTestDB(project_name, token_name='unittest'):
try:
# get the objects
tk = NDToken.fromName(token_name)
tk.delete()
pr = NDProject.fromName(project_name)
ds = pr.datasetcfg
# tk = Token.objects.get(token_name=token_name)
# pr = Project.objects.get(project_name=project_name)
# ds = Dataset.objects.get(dataset_name=pr.dataset_id)
# get the channel list
# channel_list = Channel.objects.filter(project_id=pr)
# get the correct object for the kvengine
# pd = NDProjectsDB.getProjDB(pr)
for ch in pr.projectChannels():
ch.delete()
# delete the channel table
# pd.deleteNDChannel(ch.channel_name)
# delete the channel
# ch.delete()
# delete the project database
# pd.deleteNDProject()
# delete the objects
pr.delete()
ds.delete()
# delete s3 bucket and dynamo table
if KV_ENGINE == REDIS:
CuboidIndexDB.deleteTable()
CuboidBucket.deleteBucket()
except Exception, e:
print(e)
raise e
def __init__(self, token_name, host_name=HOST_NAME):
"""Create the bucket and intialize values"""
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}.log'.format(token_name))
self.logger.addHandler(fh)
# setting up the project metadata
self.info_interface = InfoInterface(host_name, token_name)
# creating the resource interface to the remote server
# self.resource_interface = ResourceInterface(self.info_interface.dataset_name, self.info_interface.project_name, host_name, logger=self.logger)
# self.proj = self.resource_interface.getProject()
# create the s3 I/O and index objects
self.cuboidindex_db = CuboidIndexDB(self.info_interface.project_name)
self.cuboid_bucket = CuboidBucket(self.info_interface.project_name)
class Interface(object):
def __init__(self, dataset_name, project_name, host_name=HOST_NAME):
# self.resource_interface = ResourceInterface(dataset_name, project_name, host_name)
self.cuboid_bucket = CuboidBucket(project_name)
self.cuboidindex_db = CuboidIndexDB(project_name)
# def deleteToken(self):
# """Delete the Token"""
# self.resource_interface.deleteToken()
# print 'Delete successful for token {}'.format(self.token)
def deleteProject(self):
"""Delete the project"""
return NotImplemented
# delete the project from s3 and dynamo
# self.s3_projdb.deleteNDProject()
# deleting the meta-data via resource interface
# self.resource_interface.deleteToken()
# self.resource_interface.deleteProject()
# print 'Delete successful for project {}'.format(self.project_name)
def deleteChannel(self, channel_name):
"""Delete the channel"""
try:
for item in self.cuboidindex_db.queryChannelItems(channel_name):
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except Exception as e:
print (e)
# delete the channel from s3 and dynamo
# self.s3_projdb.deleteNDChannel(channel_name)
# deleting the meta-data via resource interface
# self.resource_interface.deleteChannel(channel_name)
# print 'Delete successful for channel {}'.format(channel_name)
def deleteResolution(self, channel_name, resolution):
"""Delete an existing resolution"""
try:
for item in self.cuboidindex_db.queryResolutionItems(channel_name, resolution):
print(item['supercuboid_key'])
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except Exception as e:
print (e)
from django.contrib.auth.models import User
from nduser.models import Project
from nduser.models import Dataset
from nduser.models import Token
from nduser.models import Channel
from ndproj.nddataset import NDDataset
from ndproj.ndproject import NDProject
from ndproj.ndchannel import NDChannel
from ndproj.ndtoken import NDToken
from ndlib.ndtype import *
from test_settings import *
if KV_ENGINE == REDIS:
from ndingest.nddynamo.cuboidindexdb import CuboidIndexDB
from ndingest.ndbucket.cuboidbucket import CuboidBucket
try:
CuboidIndexDB.deleteTable()
CuboidBucket.deleteBucket()
except Exception as e:
pass
def createTestDB ( project_name, channel_list=['unit_anno'], channel_type=ANNOTATION, channel_datatype=UINT32, public=PUBLIC_TRUE, ximagesize=10000, yimagesize=10000, zimagesize=1000, xvoxelres=4.0, yvoxelres=4.0, zvoxelres=3.0, scalingoption=ZSLICES, scalinglevels=5, readonly=READONLY_FALSE, propagate=NOT_PROPAGATED, window=[0,0], time=[0,15], default=False, nd_version=ND_VERSION, token_name='unittest', user='******', dataset_name="unittest", base_resolution=0):
"""Create a unit test data base on the specified sit and name"""
# setting s3backend to true if Redis and creating s3 bucket and dynamo table
if KV_ENGINE == REDIS:
s3backend = S3_TRUE
CuboidIndexDB.createTable()
CuboidBucket.createBucket()
else:
s3backend = S3_FALSE
class AwsInterface:
def __init__(self, token_name, host_name=HOST_NAME):
"""Create the bucket and intialize values"""
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}.log'.format(token_name))
self.logger.addHandler(fh)
# setting up the project metadata
self.info_interface = InfoInterface(host_name, token_name)
# creating the resource interface to the remote server
# self.resource_interface = ResourceInterface(self.info_interface.dataset_name, self.info_interface.project_name, host_name, logger=self.logger)
# self.proj = self.resource_interface.getProject()
# create the s3 I/O and index objects
self.cuboidindex_db = CuboidIndexDB(self.info_interface.project_name)
self.cuboid_bucket = CuboidBucket(self.info_interface.project_name)
# def setupNewProject(self):
# """Setup a new project if it does not exist"""
# self.resource_interface.createDataset()
# self.resource_interface.createProject()
# self.resource_interface.createToken()
def uploadExistingProject(self,
channel_name,
resolution,
start_values,
neariso=False):
"""Upload an existing project to S3"""
self.setupNewProject()
db = SpatialDB(self.proj)
# checking for channels
if channel_name is None:
channel_list = None
else:
channel_list = [channel_name]
# iterating over channels in a project
for ch in self.proj.projectChannels(channel_list):
# creating the channel resource
self.resource_interface.createChannel(ch.channel_name)
# ingest 1 or more resolutions based on user input
if resolution is None:
start_res = self.proj.datasetcfg.scalinglevels
stop_res = ch.resolution - 1
else:
start_res = resolution
stop_res = resolution - 1
# iterating over resolution
for cur_res in range(start_res, stop_res, -1):
# get the source database sizes
[image_size,
time_range] = self.proj.datasetcfg.dataset_dim(cur_res)
[xcubedim, ycubedim, zcubedim
] = cubedim = self.proj.datasetcfg.get_cubedim(cur_res)
offset = self.proj.datasetcfg.get_offset(cur_res)
[xsupercubedim, ysupercubedim, zsupercubedim
] = supercubedim = self.proj.datasetcfg.get_supercubedim(
cur_res)
# set the limits for iteration on the number of cubes in each dimension
xlimit = (image_size[0] - 1) / (xsupercubedim) + 1
ylimit = (image_size[1] - 1) / (ysupercubedim) + 1
zlimit = (image_size[2] - 1) / (zsupercubedim) + 1
# [xlimit, ylimit, zlimit] = limit = self.proj.datasetcfg.get_supercube_limit(cur_res)
[x_start, y_start, z_start] = map(div, start_values,
supercubedim)
for z in range(z_start, zlimit, 1):
for y in range(y_start, ylimit, 1):
for x in range(x_start, xlimit, 1):
try:
# cutout the data at the current resolution
data = db.cutout(ch, [
x * xsupercubedim, y * ysupercubedim,
z * zsupercubedim
], [
xsupercubedim, ysupercubedim, zsupercubedim
], cur_res).data
# generate the morton index
morton_index = XYZMorton([x, y, z])
self.logger.info("[{},{},{}] at res {}".format(
x * xsupercubedim, y * ysupercubedim,
z * zsupercubedim, cur_res))
# updating the index
# self.cuboidindex_db.putItem(ch.channel_name, cur_res, x, y, z, ch.time_range[0])
# inserting the cube
self.s3_io.putCube(ch,
ch.time_stamp[0],
morton_index,
cur_res,
blosc.pack_array(data),
neariso=neariso)
except Exception as e:
# checkpoint the ingest
self.logger.error(e)
self.checkpoint_ingest(ch.channel_name,
cur_res, x, y, z, e)
raise e
def uploadNewProject(self, config_file, start_values, neariso=False):
"""Upload a new project"""
# loading the config file and assdociated params and processors
config = Configuration()
config.load(json.loads(open(config_file, 'rt').read()))
config.load_plugins()
path_processor = config.path_processor_class
path_processor.setup(config.get_path_processor_params())
tile_processor = config.tile_processor_class
tile_processor.setup(config.get_tile_processor_params())
tile_params = config.get_tile_processor_params()
path_params = config.get_path_processor_params()
# creating the channel object from resource service
channel_name = config.config_data['database']['channel']
channel_datatype = self.info_interface.get_channel_datatype(
channel_name)
cur_res = tile_params['ingest_job']['resolution']
# loading all the parameters for image-sizes, tile-sizes, and iteration limits
[xsupercubedim, ysupercubedim,
zsupercubedim] = supercubedim = SUPER_CUBOID_SIZE
[x_start, x_end] = tile_params['ingest_job']['extent']['x']
[y_start, y_end] = tile_params['ingest_job']['extent']['y']
[z_start, z_end] = tile_params['ingest_job']['extent']['z']
[t_start, t_end] = tile_params['ingest_job']['extent']['t']
x_tilesz = tile_params['ingest_job']['tile_size']['x']
y_tilesz = tile_params['ingest_job']['tile_size']['y']
z_tilesz = tile_params['ingest_job']['tile_size']['z']
t_tilesz = tile_params['ingest_job']['tile_size']['t']
x_limit = (x_end - 1) / (x_tilesz) + 1
y_limit = (y_end - 1) / (y_tilesz) + 1
z_limit = (z_end - 1) / (z_tilesz) + 1
t_limit = (t_end - 1) / (t_tilesz) + 1
if start_values != [0, 0, 0]:
[x_start, y_start, z_start] = map(div, start_values,
[x_tilesz, y_tilesz, z_tilesz])
# iterate over t,z,y,x to ingest the data
for t in range(t_start, t_limit, 1):
for z in range(z_start, z_limit, zsupercubedim):
for y in range(y_start, y_limit, 1):
for x in range(x_start, x_limit, 1):
data = np.zeros([zsupercubedim, y_tilesz, x_tilesz],
dtype=ND_dtypetonp[channel_datatype])
for b in range(0, zsupercubedim, 1):
if z + b > z_end - 1:
break
# generate file name
file_name = path_processor.process(x, y, z + b, t)
# read the file, handle expection if the file is missing
try:
tile_handle = tile_processor.process(
file_name, x, y, z + b, t)
tile_handle.seek(0)
data[b, :, :] = np.asarray(
Image.open(tile_handle))
except IOError as e:
pass
# print "missing file", file_name
# iterate over the tile if it is larger then supercuboid size
for y_index in range(0, y_tilesz / ysupercubedim):
for x_index in range(0, x_tilesz / xsupercubedim):
# calculate the morton index
insert_data = data[:, y_index *
ysupercubedim:(y_index +
1) *
ysupercubedim, x_index *
xsupercubedim:(x_index +
1) *
xsupercubedim]
if np.any(insert_data):
morton_index = XYZMorton([
x_index +
(x * x_tilesz / xsupercubedim),
y_index +
(y * y_tilesz / ysupercubedim),
z / zsupercubedim
])
[s3_x, s3_y,
s3_z] = MortonXYZ(morton_index)
print "Morton Index {}".format(
morton_index)
self.logger.info("[{},{},{}]".format(
(x_index + x) * x_tilesz,
(y_index + y) * y_tilesz, z))
self.cuboidindex_db.putItem(
channel_name,
cur_res,
s3_x,
s3_y,
s3_z,
t,
neariso=neariso)
self.cuboid_bucket.putObject(
channel_name,
cur_res,
morton_index,
t,
blosc.pack_array(insert_data),
neariso=neariso)
# self.s3_io.putCube(ch, t, morton_index, cur_res, blosc.pack_array(insert_data), update=False, neariso=False)
def checkpoint_ingest(self, channel_name, resolution, x, y, z, e, time=0):
"""Checkpoint the progress to file"""
with closing(open('checkpoint_ingest.csv', 'wb')) as csv_file:
field_names = [
'project_name', 'channel_name', 'resolution', 'x', 'y', 'z',
'time', 'exception'
]
csv_writer = csv.DictWriter(csv_file,
delimiter=',',
fieldnames=field_names)
csv_writer.writeheader()
csv_writer.writerow({
'project_name': self.proj.project_name,
'channel_name': channel_name,
'resolution': resolution,
'x': x,
'y': y,
'z': z,
'time': time,
'exception': e.message
})
def load_checkpoint(self):
"""Load from a checkpoint file"""
return NotImplemented
def __init__(self, dataset_name, project_name, host_name=HOST_NAME): # self.resource_interface = ResourceInterface(dataset_name, project_name, host_name) self.cuboid_bucket = CuboidBucket(project_name) self.cuboidindex_db = CuboidIndexDB(project_name)
class AwsInterface:
def __init__(self, token_name, host_name=HOST_NAME):
"""Create the bucket and intialize values"""
# configuring the logger based on the dataset we are uploading
self.logger = logging.getLogger(token_name)
self.logger.setLevel(logging.INFO)
fh = logging.FileHandler('{}.log'.format(token_name))
self.logger.addHandler(fh)
# setting up the project metadata
self.info_interface = InfoInterface(host_name, token_name)
# creating the resource interface to the remote server
# self.resource_interface = ResourceInterface(self.info_interface.dataset_name, self.info_interface.project_name, host_name, logger=self.logger)
# self.proj = self.resource_interface.getProject()
# create the s3 I/O and index objects
self.cuboidindex_db = CuboidIndexDB(self.info_interface.project_name)
self.cuboid_bucket = CuboidBucket(self.info_interface.project_name)
# def setupNewProject(self):
# """Setup a new project if it does not exist"""
# self.resource_interface.createDataset()
# self.resource_interface.createProject()
# self.resource_interface.createToken()
def uploadExistingProject(self, channel_name, resolution, start_values, neariso=False):
"""Upload an existing project to S3"""
self.setupNewProject()
db = SpatialDB(self.proj)
# checking for channels
if channel_name is None:
channel_list = None
else:
channel_list = [channel_name]
# iterating over channels in a project
for ch in self.proj.projectChannels(channel_list):
# creating the channel resource
self.resource_interface.createChannel(ch.channel_name)
# ingest 1 or more resolutions based on user input
if resolution is None:
start_res = self.proj.datasetcfg.scalinglevels
stop_res = ch.resolution - 1
else:
start_res = resolution
stop_res = resolution - 1
# iterating over resolution
for cur_res in range(start_res, stop_res, -1):
# get the source database sizes
[image_size, time_range] = self.proj.datasetcfg.dataset_dim(cur_res)
[xcubedim, ycubedim, zcubedim] = cubedim = self.proj.datasetcfg.get_cubedim(cur_res)
offset = self.proj.datasetcfg.get_offset(cur_res)
[xsupercubedim, ysupercubedim, zsupercubedim] = supercubedim = self.proj.datasetcfg.get_supercubedim(cur_res)
# set the limits for iteration on the number of cubes in each dimension
xlimit = (image_size[0]-1) / (xsupercubedim) + 1
ylimit = (image_size[1]-1) / (ysupercubedim) + 1
zlimit = (image_size[2]-1) / (zsupercubedim) + 1
# [xlimit, ylimit, zlimit] = limit = self.proj.datasetcfg.get_supercube_limit(cur_res)
[x_start, y_start, z_start] = map(div, start_values, supercubedim)
for z in range(z_start, zlimit, 1):
for y in range(y_start, ylimit, 1):
for x in range(x_start, xlimit, 1):
try:
# cutout the data at the current resolution
data = db.cutout(ch, [x*xsupercubedim, y*ysupercubedim, z*zsupercubedim], [xsupercubedim, ysupercubedim, zsupercubedim], cur_res).data
# generate the morton index
morton_index = XYZMorton([x, y, z])
self.logger.info("[{},{},{}] at res {}".format(x*xsupercubedim, y*ysupercubedim, z*zsupercubedim, cur_res))
# updating the index
# self.cuboidindex_db.putItem(ch.channel_name, cur_res, x, y, z, ch.time_range[0])
# inserting the cube
self.s3_io.putCube(ch, ch.time_stamp[0], morton_index, cur_res, blosc.pack_array(data), neariso=neariso)
except Exception as e:
# checkpoint the ingest
self.logger.error(e)
self.checkpoint_ingest(ch.channel_name, cur_res, x, y, z, e)
raise e
def uploadNewProject(self, config_file, start_values, neariso=False):
"""Upload a new project"""
# loading the config file and assdociated params and processors
config = Configuration()
config.load(json.loads(open(config_file, 'rt').read()))
config.load_plugins()
path_processor = config.path_processor_class
path_processor.setup(config.get_path_processor_params())
tile_processor = config.tile_processor_class
tile_processor.setup(config.get_tile_processor_params())
tile_params = config.get_tile_processor_params()
path_params = config.get_path_processor_params()
# creating the channel object from resource service
channel_name = config.config_data['database']['channel']
channel_datatype = self.info_interface.get_channel_datatype(channel_name)
cur_res = tile_params['ingest_job']['resolution']
# loading all the parameters for image-sizes, tile-sizes, and iteration limits
[xsupercubedim, ysupercubedim, zsupercubedim] = supercubedim = SUPER_CUBOID_SIZE
[x_start, x_end] = tile_params['ingest_job']['extent']['x']
[y_start, y_end] = tile_params['ingest_job']['extent']['y']
[z_start, z_end] = tile_params['ingest_job']['extent']['z']
[t_start, t_end] = tile_params['ingest_job']['extent']['t']
x_tilesz = tile_params['ingest_job']['tile_size']['x']
y_tilesz = tile_params['ingest_job']['tile_size']['y']
z_tilesz = tile_params['ingest_job']['tile_size']['z']
t_tilesz = tile_params['ingest_job']['tile_size']['t']
x_limit = (x_end-1) / (x_tilesz) + 1
y_limit = (y_end-1) / (y_tilesz) + 1
z_limit = (z_end-1) / (z_tilesz) + 1
t_limit = (t_end-1) / (t_tilesz) + 1
if start_values != [0, 0, 0]:
[x_start, y_start, z_start] = map(div, start_values, [x_tilesz, y_tilesz, z_tilesz])
# iterate over t,z,y,x to ingest the data
for t in range(t_start, t_limit, 1):
for z in range(z_start, z_limit, zsupercubedim):
for y in range(y_start, y_limit, 1):
for x in range(x_start, x_limit, 1):
data = np.zeros([zsupercubedim, y_tilesz, x_tilesz], dtype=ND_dtypetonp[channel_datatype])
for b in range(0, zsupercubedim, 1):
if z + b > z_end - 1:
break
# generate file name
file_name = path_processor.process(x, y, z+b, t)
# read the file, handle expection if the file is missing
try:
tile_handle = tile_processor.process(file_name, x, y, z+b, t)
tile_handle.seek(0)
data[b,:,:] = np.asarray(Image.open(tile_handle))
except IOError as e:
pass
# print "missing file", file_name
# iterate over the tile if it is larger then supercuboid size
for y_index in range(0, y_tilesz/ysupercubedim):
for x_index in range(0, x_tilesz/xsupercubedim):
# calculate the morton index
insert_data = data[:, y_index*ysupercubedim:(y_index+1)*ysupercubedim, x_index*xsupercubedim:(x_index+1)*xsupercubedim]
if np.any(insert_data):
morton_index = XYZMorton([x_index+(x*x_tilesz/xsupercubedim), y_index+(y*y_tilesz/ysupercubedim), z/zsupercubedim])
[s3_x, s3_y, s3_z] = MortonXYZ(morton_index)
print "Morton Index {}".format(morton_index)
self.logger.info("[{},{},{}]".format((x_index+x)*x_tilesz, (y_index+y)*y_tilesz, z))
self.cuboidindex_db.putItem(channel_name, cur_res, s3_x, s3_y, s3_z, t, neariso=neariso)
self.cuboid_bucket.putObject(channel_name, cur_res, morton_index, t, blosc.pack_array(insert_data), neariso=neariso)
# self.s3_io.putCube(ch, t, morton_index, cur_res, blosc.pack_array(insert_data), update=False, neariso=False)
def checkpoint_ingest(self, channel_name, resolution, x, y, z, e, time=0):
"""Checkpoint the progress to file"""
with closing(open('checkpoint_ingest.csv', 'wb')) as csv_file:
field_names = ['project_name', 'channel_name', 'resolution', 'x', 'y', 'z', 'time', 'exception']
csv_writer = csv.DictWriter(csv_file, delimiter=',', fieldnames=field_names)
csv_writer.writeheader()
csv_writer.writerow({'project_name' : self.proj.project_name, 'channel_name' : channel_name, 'resolution' : resolution, 'x' : x, 'y' : y, 'z' : z, 'time' : time, 'exception' : e.message})
def load_checkpoint(self):
"""Load from a checkpoint file"""
return NotImplemented
class S3ProjectDB:
"""Database for the projects"""
def __init__(self, pr):
"""Create the database connection"""
self.pr = pr
# create connections for cuboid bucket and cuboid dyanmo table
self.cuboid_bucket = CuboidBucket(
self.pr.project_name, endpoint_url=ndingest_settings.S3_ENDPOINT)
self.cuboidindex_db = CuboidIndexDB(
self.pr.project_name,
endpoint_url=ndingest_settings.DYNAMO_ENDPOINT)
def __del__(self):
"""Close the database connection"""
self.close()
def close(self):
"""Close the database connection"""
pass
def newNDProject(self):
"""Create the database for a project."""
pass
def newNDChannel(self, channel_name):
"""Create the tables for a channel."""
pass
def deleteNDProject(self):
"""Delete a project in s3 and dyanmo"""
try:
for item in self.cuboidindex_db.queryProjectItems():
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error("Error in deleting S3 project {}. {}".format(
self.pr.project_name, e))
raise NDWSError("Error in deleting S3 project {}. {}".format(
self.pr.project_name, e))
def deleteNDChannel(self, channel_name):
"""Delete a channel in s3 and dynamo"""
try:
for item in self.cuboidindex_db.queryChannelItems(channel_name):
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error("Error in deleting S3 channel {}. {}".format(
channel_name, e))
raise NDWSError("Error in deleting S3 channel {}. {}".format(
channel_name, e))
def deleteNDResolution(self, channel_name, resolution):
"""Delete the resolution in s3 and dynamo"""
try:
for item in self.cuboidindex_db.queryResolutionItems(
channel_name, resolution):
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error(
"Error in deleting S3 channel resolution {},{}. {}".format(
channel_name, resolution, e))
raise NDWSError(
"Error in deleting S3 channel resolution {},{}. {}".format(
channel_name, resolution, e))
# Copyright 2014 NeuroData (http://neurodata.io)
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import os
import sys
sys.path += [os.path.abspath('../django')]
import ND.settings
os.environ['DJANGO_SETTINGS_MODULE'] = 'ND.settings'
from ndingest.nddynamo.cuboidindexdb import CuboidIndexDB
from ndingest.ndbucket.cuboidbucket import CuboidBucket
# setup cuboid indext table and bucket locally
CuboidIndexDB.createTable()
CuboidBucket.createBucket()
def __init__(self, pr): """Create the database connection""" self.pr = pr # create connections for cuboid bucket and cuboid dyanmo table self.cuboid_bucket = CuboidBucket(self.pr.project_name, endpoint_url=ndingest_settings.S3_ENDPOINT) self.cuboidindex_db = CuboidIndexDB(self.pr.project_name, endpoint_url=ndingest_settings.DYNAMO_ENDPOINT)
class S3ProjectDB:
"""Database for the projects"""
def __init__(self, pr):
"""Create the database connection"""
self.pr = pr
# create connections for cuboid bucket and cuboid dyanmo table
self.cuboid_bucket = CuboidBucket(self.pr.project_name, endpoint_url=ndingest_settings.S3_ENDPOINT)
self.cuboidindex_db = CuboidIndexDB(self.pr.project_name, endpoint_url=ndingest_settings.DYNAMO_ENDPOINT)
def __del__(self):
"""Close the database connection"""
self.close()
def close (self):
"""Close the database connection"""
pass
def newNDProject(self):
"""Create the database for a project."""
pass
def newNDChannel(self, channel_name):
"""Create the tables for a channel."""
pass
def deleteNDProject(self):
"""Delete a project in s3 and dyanmo"""
try:
for item in self.cuboidindex_db.queryProjectItems():
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error("Error in deleting S3 project {}. {}".format(self.pr.project_name, e))
raise NDWSError("Error in deleting S3 project {}. {}".format(self.pr.project_name, e))
def deleteNDChannel(self, channel_name):
"""Delete a channel in s3 and dynamo"""
try:
for item in self.cuboidindex_db.queryChannelItems(channel_name):
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error("Error in deleting S3 channel {}. {}".format(channel_name, e))
raise NDWSError("Error in deleting S3 channel {}. {}".format(channel_name, e))
def deleteNDResolution(self, channel_name, resolution):
"""Delete the resolution in s3 and dynamo"""
try:
for item in self.cuboidindex_db.queryResolutionItems(channel_name, resolution):
self.cuboid_bucket.deleteObject(item['supercuboid_key'])
self.cuboidindex_db.deleteItem(item['supercuboid_key'])
except botocore.exceptions.ClientError as e:
if e.response['Error']['Code'] == 'ResourceNotFoundException':
logger.warning("Resource was not accessible {}".format(e))
pass
else:
raise e
except Exception as e:
logger.error("Error in deleting S3 channel resolution {},{}. {}".format(channel_name, resolution, e))
raise NDWSError("Error in deleting S3 channel resolution {},{}. {}".format(channel_name, resolution, e))
class Test_CuboidIndexDB():
def setup_class(self):
"""Setup parameters"""
try:
CuboidIndexDB.createTable(endpoint_url=settings.DYNAMO_ENDPOINT)
except Exception as e:
pass
self.cuboid_index = CuboidIndexDB(
nd_proj.project_name, endpoint_url=settings.DYNAMO_ENDPOINT)
def teardown_class(self):
"""Teardown parameters"""
CuboidIndexDB.deleteTable(endpoint_url=settings.DYNAMO_ENDPOINT)
def test_putItem(self):
"""Test data insertion"""
# inserting three values for task 0, zvalues 0-2
x_value = 0
y_value = 0
for z_value in range(0, 2, 1):
self.cuboid_index.putItem(nd_proj.channel_name, nd_proj.resolution,
x_value, y_value, z_value)
# checking if the items were inserted
for z_value in range(0, 2, 1):
item_value = self.cuboid_index.getItem(nd_proj.channel_name,
nd_proj.resolution, x_value,
y_value, z_value)
assert (item_value['project_name'] == nd_proj.project_name)
assert (
item_value['channel_resolution_taskid'] == '{}&{}&{}'.format(
nd_proj.channel_name, nd_proj.resolution, 0))
# inserting two values for task 1, zvalues 0-1
for z_value in range(0, 1, 1):
self.cuboid_index.putItem(nd_proj.channel_name,
nd_proj.resolution,
x_value,
y_value,
z_value,
task_id=1)
# checking if the items were updated
for z_value in range(0, 1, 1):
item_value = self.cuboid_index.getItem(nd_proj.channel_name,
nd_proj.resolution, x_value,
y_value, z_value)
assert (item_value['project_name'] == nd_proj.project_name)
assert (
item_value['channel_resolution_taskid'] == '{}&{}&{}'.format(
nd_proj.channel_name, nd_proj.resolution, 1))
def test_queryProjectItems(self):
"""Test the query over SI"""
# inserting three values for task 0, zvalues 0-2
x_value = 0
y_value = 0
for z_value in range(0, 2, 1):
self.cuboid_index.putItem(nd_proj.channel_name, nd_proj.resolution,
x_value, y_value, z_value)
for item in self.cuboid_index.queryProjectItems():
assert (item['project_name'] == nd_proj.project_name)
for item in self.cuboid_index.queryChannelItems(nd_proj2.channel_name):
assert (item['channel_resolution_taskid'] == '{}&{}&{}'.format(
nd_proj2.channel_name, nd_proj.resolution, 0))
for item in self.cuboid_index.queryTaskItems(nd_proj.channel_name,
nd_proj.resolution, 1):
assert (item['channel_resolution_taskid'] == '{}&{}&{}'.format(
nd_proj2.channel_name, nd_proj.resolution, 0))
def test_deleteXYZ(self):
"""Test item deletion"""
x_value = 0
y_value = 0
for z_value in range(0, 2, 1):
value = self.cuboid_index.deleteXYZ(nd_proj.channel_name,
nd_proj.resolution, x_value,
y_value, z_value)
item = self.cuboid_index.getItem(nd_proj.channel_name,
nd_proj.resolution, x_value,
y_value, z_value)
assert (item == None)
def teardown_class(self):
"""Teardown parameters"""
CuboidIndexDB.deleteTable(endpoint_url=settings.DYNAMO_ENDPOINT)
from django.contrib.auth.models import User
from nduser.models import Project
from nduser.models import Dataset
from nduser.models import Token
from nduser.models import Channel
from ndproj.nddataset import NDDataset
from ndproj.ndproject import NDProject
from ndproj.ndchannel import NDChannel
from ndproj.ndtoken import NDToken
from ndlib.ndtype import *
from test_settings import *
if KV_ENGINE == REDIS:
from ndingest.nddynamo.cuboidindexdb import CuboidIndexDB
from ndingest.ndbucket.cuboidbucket import CuboidBucket
try:
CuboidIndexDB.deleteTable()
CuboidBucket.deleteBucket()
except Exception as e:
pass
def createTestDB(project_name,
channel_list=['unit_anno'],
channel_type=ANNOTATION,
channel_datatype=UINT32,
public=PUBLIC_TRUE,
ximagesize=10000,
yimagesize=10000,
zimagesize=1000,
xvoxelres=4.0,
yvoxelres=4.0,
def teardown_class(self):
"""Teardown class parameters"""
TileIndexDB.deleteTable(endpoint_url=settings.DYNAMO_ENDPOINT)
CuboidIndexDB.deleteTable(endpoint_url=settings.DYNAMO_ENDPOINT)
CleanupQueue.deleteQueue(nd_proj, endpoint_url=settings.SQS_ENDPOINT)
def createTestDB(project_name,
channel_list=['unit_anno'],
channel_type=ANNOTATION,
channel_datatype=UINT32,
public=PUBLIC_TRUE,
ximagesize=10000,
yimagesize=10000,
zimagesize=1000,
xvoxelres=4.0,
yvoxelres=4.0,
zvoxelres=3.0,
scalingoption=ZSLICES,
scalinglevels=5,
readonly=READONLY_FALSE,
propagate=NOT_PROPAGATED,
window=[0, 0],
time=[0, 15],
default=False,
nd_version=ND_VERSION,
token_name='unittest',
user='******',
dataset_name="unittest",
base_resolution=0):
"""Create a unit test data base on the specified sit and name"""
# setting s3backend to true if Redis and creating s3 bucket and dynamo table
if KV_ENGINE == REDIS:
s3backend = S3_TRUE
CuboidIndexDB.createTable()
CuboidBucket.createBucket()
else:
s3backend = S3_FALSE
unituser = User.objects.get(username=user)
ds = NDDataset(
Dataset(dataset_name=dataset_name,
user=unituser,
ximagesize=ximagesize,
yimagesize=yimagesize,
zimagesize=zimagesize,
xoffset=0,
yoffset=0,
zoffset=1,
xvoxelres=xvoxelres,
yvoxelres=yvoxelres,
zvoxelres=zvoxelres,
scalingoption=scalingoption,
scalinglevels=scalinglevels,
public=PUBLIC_TRUE,
dataset_description="Unit test"))
ds.create()
# make the project entry
pr = NDProject(
Project(project_name=project_name,
project_description='Unit test',
user=unituser,
dataset=ds._ds,
nd_version=nd_version,
host='localhost',
kvengine=KV_ENGINE,
kvserver=KV_SERVER,
s3backend=s3backend))
pr.create()
# create a token
tk = NDToken(
Token(token_name=token_name,
user=unituser,
token_description='Unit test token',
project_id=pr.project_name,
public=public))
tk.create()
# get the correct object for the kvengine
# pd = NDProjectsDB.getProjDB(NDProjectpr)
# create the database
# pd.newNDProject()
try:
for channel_name in channel_list:
ch = NDChannel(
Channel(channel_name=channel_name,
channel_type=channel_type,
channel_datatype=channel_datatype,
channel_description='Unit test channel',
project_id=pr.project_name,
readonly=readonly,
propagate=propagate,
resolution=base_resolution,
exceptions=1,
starttime=time[0],
endtime=time[1],
startwindow=window[0],
endwindow=window[1],
default=default))
# create a channel
ch.create()
# create the channel table
# pd.newNDChannel(ch.channel_name)
except Exception, e:
print(e)
raise e