|
| 1 | +# Copyright 2016 Google Inc. |
| 2 | +# |
| 3 | +# Licensed under the Apache License, Version 2.0 (the "License"); |
| 4 | +# you may not use this file except in compliance with the License. |
| 5 | +# You may obtain a copy of the License at |
| 6 | +# |
| 7 | +# |
| 8 | +# Unless required by applicable law or agreed to in writing, software |
| 9 | +# distributed under the License is distributed on an "AS IS" BASIS, |
| 10 | +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 11 | +# See the License for the specific language governing permissions and |
| 12 | +# limitations under the License. |
| 13 | + |
| 14 | +"""Demonstrates how to connect to Cloud Bigtable and run some basic operations. |
| 15 | +# http://www.apache.org/licenses/LICENSE-2.0 |
| 16 | +Prerequisites: |
| 17 | +- Create a Cloud Bigtable cluster. |
| 18 | + https://cloud.google.com/bigtable/docs/creating-cluster |
| 19 | +- Set your Google Application Default Credentials. |
| 20 | + https://developers.google.com/identity/protocols/application-default-credentials |
| 21 | +""" |
| 22 | + |
| 23 | +import datetime |
| 24 | +import argparse |
| 25 | +from google.cloud import bigtable |
| 26 | + |
| 27 | +def run_table_operations(project_id, instance_id, table_id): |
| 28 | + ''' Create bigtable and perform different operations on it. |
| 29 | + :project_id: Project id of the client |
| 30 | + :instance_id: Instance of the client |
| 31 | + :table_id: Table id to create table. |
| 32 | + ''' |
| 33 | + |
| 34 | + client = bigtable.Client(project=project_id, admin=True) |
| 35 | + instance = client.instance(instance_id) |
| 36 | + table = instance.table(table_id) |
| 37 | + |
| 38 | + table_exists = exists() |
| 39 | + print('Checking if table {} exists...'.format(table_id)) |
| 40 | + try: |
| 41 | + |
| 42 | + if table_exists: |
| 43 | + print('Table {} already exists.'.format(table_id)) |
| 44 | + except : |
| 45 | + print('Error checking table exists.'.format(table_id)) |
| 46 | + |
| 47 | + if not table_exists: |
| 48 | + print('Creating table {}.'.format(table_id)) |
| 49 | + try: |
| 50 | + table.create() |
| 51 | + except : |
| 52 | + print('Error creating table {}.'.format(table_id)) |
| 53 | + else: |
| 54 | + print('Created table {}.'.format(table_id)) |
| 55 | + |
| 56 | + try: |
| 57 | + tables = instance.list_tables() #table.get_tables() |
| 58 | + |
| 59 | + except: |
| 60 | + print('Error listing tables.') |
| 61 | + |
| 62 | + print('Listing tables in current project...') |
| 63 | + if tables != []: |
| 64 | + for t in tables: |
| 65 | + print t.name |
| 66 | + else: |
| 67 | + print('No table exists in current project...') |
| 68 | + |
| 69 | + |
| 70 | + print('Printing table metadata...') |
| 71 | + try: |
| 72 | + print(table.name) |
| 73 | + except : |
| 74 | + print('Error retrieving table metadata.') |
| 75 | + |
| 76 | + return 1 |
| 77 | + print('Creating column family cf1 with with MaxAge GC Rule...') |
| 78 | + # [START bigtable_create_family_gc_max_age] |
| 79 | + # Create a column family with GC policy : maximum age |
| 80 | + # where age = current time minus cell timestamp |
| 81 | + column_family_id1 = 'cf1' |
| 82 | + try: |
| 83 | + # Define the GC rule to retain data with max age of 5 days |
| 84 | + max_age = datetime.timedelta(days=5) |
| 85 | + max_age_rule = table.max_age_gc_rule(max_age) |
| 86 | + print('Created column MaxAge GC rule') |
| 87 | + except: |
| 88 | + print('Error creating MaxAge GC Rule.') |
| 89 | + |
| 90 | + try: |
| 91 | + cf1 = table.column_family(column_family_id1, max_age_rule) |
| 92 | + cf1.create() |
| 93 | + print('Created column family cf1 with MaxAge GC Rule.') |
| 94 | + except: |
| 95 | + print('Error creating column family with MaxAge GC Rule.') |
| 96 | + # [End bigtable_create_family_gc_max_age] |
| 97 | + |
| 98 | + |
| 99 | + print('Creating column family cf2 with max versions GC rule...') |
| 100 | + # [START bigtable_create_family_gc_max_versions] |
| 101 | + # Create a column family with GC policy : most recent N versions |
| 102 | + # where 1 = most recent version |
| 103 | + column_family_id2 = 'cf2' |
| 104 | + try: |
| 105 | + # Define the GC policy to retain only the most recent 2 versions |
| 106 | + max_versions = 2 |
| 107 | + max_versions_rule = table.max_versions_gc_rule(max_versions) |
| 108 | + print('Created column Max Versions GC Rule.') |
| 109 | + except: |
| 110 | + print('Error creating Max Versions GC Rule.') |
| 111 | + |
| 112 | + try: |
| 113 | + cf2 = table.column_family(column_family_id2, max_versions_rule) |
| 114 | + cf2 .create() |
| 115 | + print('Created column family cf2 with Max Versions GC Rule.') |
| 116 | + except: |
| 117 | + print('Error creating column family with Max Versions GC Rule.') |
| 118 | + # [END bigtable_create_family_gc_max_versions] |
| 119 | + |
| 120 | + |
| 121 | + print('Creating column family cf3 with union GC rule...') |
| 122 | + # [START bigtable_create_family_gc_union] |
| 123 | + # Create a column family with GC policy to drop data that matches at least one condition. |
| 124 | + column_family_id3 = 'cf3' |
| 125 | + try: |
| 126 | + # GC rule : Drop max age rule OR the most recent version rule. |
| 127 | + union = [max_age_rule, max_versions_rule] |
| 128 | + union_rule = table.gc_rule_union(union) |
| 129 | + print('Created column Union GC Rule.') |
| 130 | + except: |
| 131 | + print('Error creating Union GC Rule.') |
| 132 | + |
| 133 | + try: |
| 134 | + cf3 = table.column_family(column_family_id3, union_rule) |
| 135 | + cf3.create() |
| 136 | + print('Created column family cf3 with Union GC rule') |
| 137 | + except: |
| 138 | + print('Error creating column family with Union GC rule') |
| 139 | + # [END bigtable_create_family_gc_union] |
| 140 | + |
| 141 | + |
| 142 | + |
| 143 | + print('Creating column family cf4 with Intersection GC rule...') |
| 144 | + # [START bigtable_create_family_gc_intersection] |
| 145 | + # Create a column family with GC policy to drop data that matches all conditions |
| 146 | + column_family_id4 = 'cf4' |
| 147 | + |
| 148 | + try: |
| 149 | + # GC rule: Drop cells older than 5 days AND older than the most recent 2 versions |
| 150 | + intersection = [max_age_rule, max_versions_rule] |
| 151 | + intersection_rule = table.gc_rule_intersection(union) |
| 152 | + print('Created column Intersection GC Rule.') |
| 153 | + except: |
| 154 | + print('Error creating Intersection GC Rule.') |
| 155 | + |
| 156 | + try: |
| 157 | + cf4 = table.column_family(column_family_id4, intersection_rule) |
| 158 | + cf4.create() |
| 159 | + print('Created column family cf4 with Intersection GC rule.') |
| 160 | + except: |
| 161 | + print('Error creating column family with Intersection GC rule.') |
| 162 | + # [END bigtable_create_family_gc_intersection] |
| 163 | + |
| 164 | + |
| 165 | + print('Creating column family cf5 with a Nested GC rule...') |
| 166 | + # [START bigtable_create_family_gc_nested] |
| 167 | + # Create a column family with nested GC policys. |
| 168 | + |
| 169 | + # Create a nested GC rule: |
| 170 | + # Drop cells that are either older than the 10 recent versions |
| 171 | + # OR |
| 172 | + # Drop cells that are older than a month AND older than the 2 recent versions |
| 173 | + column_family_id5 = 'cf5' |
| 174 | + try: |
| 175 | + # Drop cells that are either older than the 10 recent versions |
| 176 | + max_versions_rule1 = table.max_versions_gc_rule(10) |
| 177 | + |
| 178 | + # Drop cells that are older than a month AND older than the 2 recent versions |
| 179 | + max_age = datetime.timedelta(days=30) |
| 180 | + max_age_rule = table.max_age_gc_rule(max_age) |
| 181 | + max_versions_rule2 = table.max_versions_gc_rule(2) |
| 182 | + intersection = [max_age_rule, max_versions_rule2] |
| 183 | + intersection_rule = table.gc_rule_intersection(intersection) |
| 184 | + |
| 185 | + # This nesting is done with union rule since it is OR between the selected rules. |
| 186 | + nest = [max_versions_rule1, intersection_rule] |
| 187 | + nested_rule = table.gc_rule_union(nest) |
| 188 | + print('Created column Nested GC Rule.') |
| 189 | + except: |
| 190 | + print('Error creating Nested GC Rule.') |
| 191 | + |
| 192 | + try: |
| 193 | + cf5 = table.column_family(column_family_id5, nested_rule) |
| 194 | + cf5.create() |
| 195 | + print('Created column family cf5 with a Nested GC rule.') |
| 196 | + except: |
| 197 | + print('Error creating column family with a Nested GC rule.') |
| 198 | + # [END bigtable_create_family_gc_nested] |
| 199 | + |
| 200 | + |
| 201 | + print('Printing ID and GC Rule for all column families...') |
| 202 | + cf = table.list_column_families() |
| 203 | + column_families = [key for key,val in cf.items()] |
| 204 | + for cf,gc_rule in sorted(cf.items()): |
| 205 | + print "Column Family:", cf |
| 206 | + print "GC Rule:" |
| 207 | + print list_gc_rules(gc_rule) |
| 208 | + |
| 209 | +def list_gc_rules(column_family): |
| 210 | + return column_family.gc_rule.to_pb() |
| 211 | + |
| 212 | +def exists(): |
| 213 | + return False |
| 214 | + |
| 215 | +def delete_table(project_id, instance_id, table_id): |
| 216 | + ''' Delete bigtable. |
| 217 | + :project_id: Project id of the client |
| 218 | + :instance_id: Instance of the client |
| 219 | + :table_id: Table id to create table. |
| 220 | + ''' |
| 221 | + |
| 222 | + client = bigtable.Client(project=project_id, admin=True) |
| 223 | + instance = client.instance(instance_id) |
| 224 | + table = instance.table(table_id) |
| 225 | + table_exists = False |
| 226 | + |
| 227 | + print('Checking if {} table exists...'.format(table_id)) |
| 228 | + try: |
| 229 | + table_exists = table.exists() |
| 230 | + except : |
| 231 | + print('Error checking table exists.'.format(table_id)) |
| 232 | + |
| 233 | + #if not table_exists: |
| 234 | + if False: |
| 235 | + print('Table {} does not exists.'.format(table_id)) |
| 236 | + else: |
| 237 | + try: |
| 238 | + print('Deleting {} table.'.format(table_id)) |
| 239 | + table.delete() |
| 240 | + except: |
| 241 | + print('Error deleting {} table.'.format(table_id)) |
| 242 | + else: |
| 243 | + print('Deleted {} table.'.format(table_id)) |
| 244 | + |
| 245 | + |
| 246 | +if __name__ == '__main__': |
| 247 | + parser = argparse.ArgumentParser( |
| 248 | + description=__doc__, |
| 249 | + formatter_class=argparse.ArgumentDefaultsHelpFormatter) |
| 250 | + |
| 251 | + parser.add_argument('command', |
| 252 | + help='run or delete. Operation to perform on table.') |
| 253 | + parser.add_argument( |
| 254 | + '--table', |
| 255 | + help='Cloud Bigtable Table name.', |
| 256 | + default='Hello-Bigtable') |
| 257 | + |
| 258 | + parser.add_argument('project_id', help='Your Cloud Platform project ID.') |
| 259 | + parser.add_argument( |
| 260 | + 'instance_id', help='ID of the Cloud Bigtable instance to connect to.') |
| 261 | + |
| 262 | + args = parser.parse_args() |
| 263 | + |
| 264 | + if args.command.lower() == 'run': |
| 265 | + run_table_operations(args.project_id, args.instance_id, args.table) |
| 266 | + delete_table(args.project_id, args.instance_id, args.table) |
| 267 | + elif args.command.lower() == 'delete': |
| 268 | + delete_table(args.project_id, args.instance_id, args.table) |
| 269 | + else: |
| 270 | + print ('Command should be either run or delete.\n Use argument -h, --help to show help and exit.') |
| 271 | + |
| 272 | + |
0 commit comments