This document is optimized for LLM and AI agent parsing. Uses structured key-value pairs, hierarchical organization, and semantic chunking. Not for human readability - designed for efficient information extraction by AI systems. # TIGERGRAPH_CLOUD_CLASSIC_AGENT_REFERENCE ## METADATA product: TigerGraph_Cloud_Classic version: 3.x last_updated: 2026_04_16 target_audience: ai_agents_llms format: structured_semantic_chunks parsing_strategy: extract_by_section_headings status: legacy_deprecated_for_new_users successor: TigerGraph_Savanna ## TIER_1_CRITICAL_INFORMATION ### DEPRECATION_NOTICE status: no_longer_available_for_new_users successor: TigerGraph_Savanna successor_url: https://tgcloud.io migration_recommendation: new_projects_should_use_savanna existing_users: continue_access_at_classic_tgcloud_io ### SUPPORT_MATRIX support_primary: channel: discord url: https://discord.gg/tnhpgR3j response_guarantee: immediate_community_support message_template: having_issues_reach_out_to_discord_team_tigergraph_will_help_for_sure support_secondary: channel: developer_forum url: https://dev.tigergraph.com/forum type: asynchronous_qa support_email: address: sales@tigergraph.com type: enterprise_sales_support support_tickets: availability: paid_tier_solutions_only free_tier: answered_when_bandwidth_allows ### PRODUCT_OVERVIEW product: name: TigerGraph_Cloud_Classic type: fully_managed_cloud_database engine: tigergraph_enterprise_server deployment_model: instance_based_clusters architecture: integrated_storage_and_compute url: https://classic.tgcloud.io suite_includes: - admin_portal - graphstudio - insights cloud_providers: - aws_amazon_web_services - gcp_google_cloud_platform - azure_microsoft_azure regions: total: 20_plus_across_all_providers aws: multiple_regions_available gcp: multiple_regions_available azure: multiple_regions_available note: check_cloud_portal_for_current_region_list pricing: instance_based multigraph: supported_v3_0_5_plus ## GETTING_STARTED ### ACCOUNT_SETUP url: https://classic.tgcloud.io steps: - step_1: sign_up_with_company_name_personal_name_email - step_2: verify_email_check_spam_folder - step_3: log_in_to_cloud_portal - step_4: create_cluster - step_5: access_graphstudio_via_tools_dropdown ### CLUSTER_PROVISIONING process: - select_cloud_provider_aws_gcp_azure - choose_region - select_instance_type_including_free_tier - optionally_set_replication_factor_2_or_3_for_ha - provision_and_wait_for_ready_status ha_configuration: replication_factor: 2_or_3 minimum_instances: 3 mode: active_active formula: total_instances_equals_replication_x_partitioning ### STARTER_KITS description: prebuilt_graph_solutions_with_schema_data_queries post_provisioning_steps: - step_1: connect_to_graphstudio_via_open_solution_via_domain - step_2: go_to_load_data_tab_click_load_button - step_3: go_to_write_query_tab_click_install_button - step_4: run_queries_on_sample_data overview_videos: https://www.tigergraph.com/starterkits/ ## TIER_QUOTAS_AND_PRICING ### FREE_TIER cost: free free_credit: 25_usd credit_validity: 30_days_after_first_use instances_allowed: 1_free_tier_per_account backup: not_included support: not_included auto_stop: after_1_hour_inactivity auto_terminate: after_7_days_inactivity data_preservation: data_preserved_on_stop_if_credits_remain data_deletion: solutions_deleted_when_no_credits_and_no_credit_card ### BASIC_ACCOUNT_WITH_CREDIT_CARD vcpu_limit: 160 instance_limit: 20 backup_restore: included_for_non_free_tier_instances support: included ### FREE_CREDIT_ACCOUNT_NO_CARD vcpu_limit: 16 solution_limit: 2 free_tier_limit: 1_per_account ## CORE_TECHNOLOGY_STACK ### GRAPH_DATA_MODEL vertex: definition: entity_node_with_unique_id properties: name_type_attributes indexing: primary_index_automatic_plus_optional_secondary example: person_vertex_with_user_id_name_age_email edge: definition: relationship_between_two_vertices direction: from_source_to_target properties: optional_weight_timestamp_metadata example: knows_edge_from_person_to_person_with_since_property property_types: numeric: [int, uint, float, double] text: [string] temporal: [datetime] boolean: [bool] collections: [list, map] note: extended_int_types_uint8_uint16_etc_are_4x_only indexing: primary_index: vertex_id_automatic secondary_index: optional_on_any_property composite_index: multiple_properties_together purpose: accelerate_where_clauses_and_filtering ### QUERY_LANGUAGE_GSQL name: GSQL type: declarative_graph_query_language designed_for: graph_pattern_matching_traversal syntax: c_like_with_graph_extensions additional_languages: none_gsql_only core_concepts: pattern_matching: select_from_vertex_edges_target_vertex_syntax multi_hop: traverse_multiple_edges_in_single_query accumulators: aggregate_values_during_traversal filtering: where_clauses_at_each_step results: print_or_return_computed_results example_query_structure: opening: "CREATE QUERY query_name(PARAMS) FOR GRAPH graph_name {" variables: "SumAccum @counter;" initialization: "start = {vertex_id};" pattern: "result = SELECT t FROM start:s -(edge:e)-> target:t WHERE ...;" output: "PRINT result;" closing: "}" ## API_REFERENCE ### REST_API_ENDPOINTS base_urls: cloud_classic: https://cluster_domain.i.tgcloud.io:9000 authentication: bearer_token_required note: cluster_specific_url_from_cloud_portal endpoints: query_execution: path: /query/{graph_name}/{query_name} method: [get, post] purpose: run_installed_gsql_query params: optional_query_parameters example: /query/social/bfs_neighbors?source=user_1 vertex_crud: insert: path: /graph/{graph}/vertices/{type} method: post body: json_properties read: path: /graph/{graph}/vertices/{type}/{id} method: get update: path: /graph/{graph}/vertices/{type}/{id} method: put body: json_updated_properties delete: path: /graph/{graph}/vertices/{type}/{id} method: delete edge_crud: insert: path: /graph/{graph}/edges/{type} method: post body: json_from_to_properties read: path: /graph/{graph}/edges/{type} method: get params: from_id_to_id_optional delete: path: /graph/{graph}/edges/{type} method: delete params: from_id_to_id_required graph_info: stats: path: /graph/{graph}/stats method: get returns: vertex_edge_counts schema: path: /graph/{graph}/schema method: get returns: full_schema_definition vertices: path: /graph/{graph}/vertices method: get returns: vertex_type_definitions edges: path: /graph/{graph}/edges method: get returns: edge_type_definitions endpoints: path: /gsqlserver/endpoints method: get returns: list_installed_queries authentication: method: bearer_token obtain_token: path: /api/token method: post body: username_password usage: Authorization_header_Bearer_token ### PYTHON_SDK_PYTIGERGRAPH package_name: pyTigerGraph installation: pip_install_pyTigerGraph python_version: 3.7_plus class_name: TigerGraphConnection initialization: host: cluster_domain.i.tgcloud.io graphname: graph_name username: tigergraph password: password rest_port: 9000 ssl: true methods_query: runQuery: execute_installed_query_with_params gsql: execute_adhoc_gsql methods_vertex: upsertVertex: insert_or_update_vertex getVertex: retrieve_single_vertex getVertices: retrieve_multiple_vertices_limit deleteVertex: remove_vertex methods_edge: upsertEdge: insert_or_update_edge getEdges: retrieve_edges_from_vertex methods_graph: getGraphStats: vertex_edge_counts getSchema: full_schema getVertexTypes: list_vertex_types getEdgeTypes: list_edge_types ## TOOLS_AND_INTERFACES ### GRAPHSTUDIO type: visual_ide_interface_primary_ui access: via_cloud_portal_tools_dropdown_open_solution_via_domain included: all_tiers features: - schema_design_drag_drop_editor - schema_mapping_creation - data_loading_ui_with_load_button - query_builder_write_query_tab - query_installation_install_button - result_visualization_graph_table_chart - data_import_local_file_upload - multigraph_support_v3_0_5_plus - team_collaboration_shared_workspaces ### GSQL_WEB_SHELL type: interactive_command_line_gsql access: via_cloud_portal_solution_panel availability: paid_tiers_only features: - gsql_command_execution - data_definition_and_loading - s3_loader_parquet_format - kafka_loader - complex_loading_jobs ### ADMIN_PORTAL access: via_cloud_portal features: - performance_monitoring_per_instance - backup_restore_non_free_tier - user_access_management - log_viewing_debugging - license_version_info - system_configuration ### TIGERGRAPH_INSIGHTS type: no_code_dashboard_builder included: part_of_cloud_classic_suite features: - drag_drop_widget_builder - prebuilt_visualization_templates - real_time_metric_updates - drill_down_filters - custom_parameters - pdf_export - sharing_access_control ## ALGORITHMS_LIBRARY algorithms_count: 50_plus categories: - centrality - community_detection - pathfinding - similarity - node_ranking - traversal centrality_algorithms: pagerank: node_importance_ranking betweenness: node_bridge_importance closeness: node_proximity_measure degree: direct_connection_count community_detection: louvain: modularity_optimization weakly_connected_components: reachability_groups strongly_connected_components: cycles pathfinding: shortest_path: optimal_route_minimal_cost all_pairs_shortest_path: distance_matrix bfs: breadth_first_traversal dfs: depth_first_traversal similarity: jaccard: set_similarity cosine: vector_similarity euclidean: distance_metric ranking: topic_sensitive_pagerank: personalized_importance eigenvector_centrality: recursive_importance traversal: reachability: all_reachable_nodes k_hop_neighbors: nodes_within_distance ## DATA_LOADING_INGESTION ### SOURCE_CONNECTORS supported_sources: - local_file_upload_via_graphstudio - amazon_s3_buckets - apache_kafka_streams - apache_spark_jdbc_driver - http_endpoints s3_and_kafka_note: available_via_gsql_web_shell_paid_tiers_only spark_driver: https://github.com/tigergraph/ecosys/tree/master/tools/etl ### LOADING_METHOD_1_GRAPHSTUDIO_UI platform: cloud_classic_all_tiers complexity: very_easy steps: - navigate_to_load_data_tab_in_graphstudio - upload_local_file_or_configure_source - map_columns_to_schema - click_load_button - monitor_progress ### LOADING_METHOD_2_GSQL_WEB_SHELL platform: cloud_classic_paid_tiers_only complexity: medium repeatable: yes scriptable: yes structure: create: CREATE_LOADING_JOB_name_FOR_GRAPH_graph define: DEFINE_FILENAME_or_KAFKA_TOPIC load: LOAD_source_TO_VERTEX_type_VALUES run: RUN_LOADING_JOB_with_using_params token_functions: string: split_substr_upper_lower_trim_replace conversion: to_int_to_float_to_bool_to_datetime hashing: md5_sha1_sha256 temporal: now_unix_date_date conditional: if_coalesce math: abs_round ### LOADING_METHOD_3_KAFKA_STREAMING type: real_time_continuous platform: paid_tiers_via_gsql_web_shell format: json_messages_in_topics ## CLUSTER_MANAGEMENT ### STOPPING_VS_TERMINATING stop: action: pauses_cluster data: preserved_on_disk billing: no_compute_charges restart: data_available_after_restart note: free_tier_auto_stops_after_1_hour_inactivity terminate: action: deletes_cluster_permanently data: not_preserved billing: all_charges_stop recovery: not_possible_unless_backup_exists ### BACKUP_AND_RESTORE availability: paid_tiers_with_credit_card_only access: admin_portal note: solutions_provisioned_before_credit_card_entry_do_not_get_backup ### DISTRIBUTED_CONFIGURATION partitioning: automatic_data_distribution replication: active_active_all_copies_serve_queries ha_minimum: 3_instances_for_replicated_system formula: instances_equals_replication_factor_x_partitioning_factor ## PERFORMANCE_CHARACTERISTICS latency: single_hop_query: sub_millisecond multi_hop_query: sub_second billion_vertex_graph: sub_second throughput: vertices_per_second: millions edges_per_second: millions scalability: horizontal: distribute_across_cluster_nodes vertical: increase_instance_size optimization_strategies: indexing: secondary_indexes_on_filtered_properties query_design: early_filtering_limit_reduce_results schema_design: minimize_edge_properties instance_sizing: choose_adequate_vcpu_and_ram ## SCHEMA_DESIGN_BEST_PRACTICES vertex_id_design: meaningful: use_natural_ids_not_surrogate immutable: never_change_after_creation domain_context: include_type_prefix partitioning: consider_distribution_hints property_selection: inclusion: only_what_you_query typing: use_correct_data_type denormalization: ok_for_read_heavy timestamps: include_for_auditing edge_design: minimal_properties: keep_lightweight reverse_edges: enable_bidirectional_traversal weighting: include_for_algorithms temporal: timestamps_for_analysis indexing_strategy: primary: automatic_on_vertex_id secondary: on_filtered_properties composite: for_multi_property_queries balance: tradeoff_read_speed_vs_write_cost ## SECURITY_FEATURES authentication: local_users: username_password ldap: enterprise_directory saml_oauth: identity_providers api_tokens: programmatic_access authorization: rbac: role_based_access_control row_level: vertex_edge_filtering query_level: restrict_query_execution encryption: at_rest: encrypted in_transit: encrypted enforcement: configured_by_default audit: logging: all_operations_tracked compliance: hipaa_pci_dss_soc2 retention: configurable_policies ## TROUBLESHOOTING_DECISION_TREE issue_cannot_find_activation_email: solution: check_spam_folder_resend_verification escalation: contact_support_at_tigergraph_for_manual_verification issue_free_tier_auto_stopped: solution: manually_restart_from_cloud_portal cause: no_user_activity_for_1_hour issue_free_tier_terminated: solution: reprovision_new_instance cause: 7_days_of_inactivity issue_data_lost_after_terminate: solution: data_cannot_be_recovered prevention: use_backup_restore_on_paid_tiers issue_slow_queries: solution: add_indexes_use_explain_plan_optimize_filters alternative: upgrade_instance_type_for_more_resources issue_connection_refused: solution: verify_cluster_is_running_check_status_in_portal common_cause: cluster_auto_stopped_due_to_inactivity issue_gsql_web_shell_unavailable: solution: upgrade_to_paid_tier cause: gsql_web_shell_only_available_on_paid_tiers issue_need_help: immediate: discord_https_discord_gg_tnhpgR3j message: tigergraph_team_will_help_for_sure forum: https_dev_tigergraph_com_forum ## COMPARISON_WITH_SAVANNA ### KEY_DIFFERENCES architecture: classic: integrated_storage_and_compute savanna: separate_storage_and_compute cloud_providers: classic: aws_gcp_azure_all_three savanna: aws_primary_expanding_to_gcp_azure url: classic: https://classic.tgcloud.io savanna: https://tgcloud.io version: classic: 3_x_releases savanna: 4_x_latest_releases primary_ui: classic: graphstudio_full_visual_ide savanna: gsql_editor_plus_graph_explorer_graphstudio_as_addon deployment_model: classic: cluster_based_solutions savanna: workspace_based_workgroups monitoring: classic: admin_portal_only savanna: built_in_grafana_dashboard_plus_admin_portal_plus_alerting query_languages: classic: gsql_only savanna: gsql_plus_opencypher_plus_iso_gql data_loading: classic: graphstudio_ui_upload_s3_kafka_via_gsql_web_shell_paid savanna: load_data_ui_local_s3_gcs_abs solution_kits: classic: starter_kits savanna: 9_solution_kits_with_guided_ui pricing: classic: instance_based savanna: usage_based_tgsize_tiers free_tier: classic: 25_usd_credit_30_days savanna: free_credits_valid_1_year ### MIGRATION_NOTE recommendation: new_projects_should_use_savanna classic_status: legacy_existing_users_only ## QUICK_REFERENCE ### CLOUD_PORTAL_URLS login: https://classic.tgcloud.io graphstudio_access: cloud_portal_tools_dropdown_open_solution_via_domain gsql_web_shell: cloud_portal_solution_panel_paid_tiers ### PYTHON_SDK_SNIPPETS import: from pytigergraph import TigerGraphConnection connect: conn = TigerGraphConnection(host="cluster_domain.i.tgcloud.io", graphname=..., username=..., password=..., sslPort=443, tgCloud=True) run_query: conn.runQuery(query_name, params) insert_vertex: conn.upsertVertex(type, id, attributes) insert_edge: conn.upsertEdge(from_type, from_id, edge_type, to_type, to_id, attributes) get_stats: conn.getGraphStats() ### REST_API_CURL_EXAMPLES run_query: "curl -H 'Authorization: Bearer TOKEN' https://cluster.i.tgcloud.io:9000/query/graph/query_name?param=value" insert_vertex: "curl -X POST -H 'Authorization: Bearer TOKEN' https://cluster.i.tgcloud.io:9000/graph/social/vertices/person -H 'Content-Type: application/json' -d {}" get_vertex: "curl -H 'Authorization: Bearer TOKEN' https://cluster.i.tgcloud.io:9000/graph/social/vertices/person/id" get_stats: "curl -H 'Authorization: Bearer TOKEN' https://cluster.i.tgcloud.io:9000/graph/social/stats" ## RESOURCE_URLS_ACTIONABLE cloud_classic_portal: https://classic.tgcloud.io cloud_classic_docs: https://docs.tigergraph.com/cloud/main/start/overview cloud_classic_faq: https://docs.tigergraph.com/cloud/main/support/faqs comparison_table: https://docs.tigergraph.com/savanna/main/overview/comparison_table graphstudio_docs: https://docs.tigergraph.com/gui/3.10/graphstudio/overview gsql_docs: https://docs.tigergraph.com/gsql/current/ algorithms_docs: https://docs.tigergraph.com/graph-algorithm-library/current/ starter_kits: https://www.tigergraph.com/starterkits/ discord_support: https://discord.gg/tnhpgR3j forum_support: https://dev.tigergraph.com/forum github_ecosystem: https://github.com/tigergraph/ecosys sales_contact: sales@tigergraph.com savanna_migration: https://tgcloud.io ## AGENT_PARSING_INSTRUCTIONS how_to_extract: - identify_section_by_heading - extract_key_value_pairs_directly - parse_nested_structures_hierarchically - follow_decision_trees_for_troubleshooting - reference_urls_for_external_resources priority_sections: 1_critical: deprecation_notice_support_information_getting_started 2_essential: apis_sdks_tools_graphstudio 3_important: data_loading_schema_design_cluster_management 4_reference: algorithms_security_comparison_with_savanna extraction_targets: - product_capabilities_and_limitations - setup_and_access_steps - api_endpoints - error_solutions - resource_urls - code_examples - classic_vs_savanna_differences ## END_DOCUMENT total_sections: 25_plus structured_for: ai_agent_comprehension not_for: human_reading last_updated: 2026_04_16_ist