Destinado a administradores de banco de dados como recomendações para estabelecer os espaços de trabalho do produto em um Enterprise Geodatabase (OracleAE, SQL ServerAE, PostgreSQLAE).
_______________________________________________________________
Índice
Esri MCSExtensões Esri MCS ArcGISPráticas recomendadas para bancos de dados empresariais MCSFAQ - Perguntas FrequentesGuias ArcGIS ProGuias ArcMap 10.8.xGuias versões anterioresScripts modelo para bancos de dados empresariais MCSGeodatabasesScripts modelo de banco de dadosPráticas recomendadas para conexões de banco de dadosLinks para documentação de banco de dadosPráticas recomendadas para armazenamento de banco de dadosPor que usar múltiplos tablespaces e RAID 10BAARF - Batalha Contra Qualquer RAID 5Configurando o servidor Oracle para VLDB - Cary Millsap, OracleConfiguração ótima de armazenamento facilitada - Juan Loaiza, Oracle
________________________________________________________________________________________________
Práticas recomendadas para bancos de dados empresariais MCS - versões anteriores
________________________________________________________________________________________________
Esri MCS
Mapping and Charting Solutions (MCS) da Esri é usado por todos os tipos de organizações para melhorar fluxos de trabalho e atender às necessidades mais desafiadoras de produção cartográfica e mapeamento.
Sites principais de referência |Wiki -Estratégias de design do sistema | community.esri.com - Estratégias de design do sistema
Uma referência técnica Esri AE fevereiro 2026
________________________________________________________________________________________________
Extensões Esri MCS ArcGIS
________________________________________________________________________________________________
Práticas recomendadas para bancos de dados empresariais MCS
Os guias são destinados a administradores de banco de dados como recomendações para estabelecer os espaços de trabalho do produto em um Enterprise Geodatabase (OracleAE, SQL ServerAE, PostgreSQLAE).
* uma implantação bem-sucedida da aplicação empresarial ArcGIS começa com um design físico forte do banco de dados e as melhores práticas de administração do banco de dados.
Crie um geodatabase para cada espaço de trabalho do produto software.
00pm - banco de dados de produção Production Mapping & Defense Mapping
Mapeamento topográfico - MGCP, TDS, etc.
seus dados GIS - crie geodatabase separado para cada
escala cartográfica ex: pm50 (1:50k); pm250 (1:250k)
pl - espaço de trabalho da biblioteca do produto - Product Library* (* apenas ArcMap)
rev - espaço de trabalho data reviewer - Data Reviewer
wmx - espaço de trabalho workflow manager - Workflow Manager
mar1 - banco de dados cartográfico marítimo - Náutico
seus dados GIS - crie geodatabase separado para cada
escala cartográfica ex: mar500k (1:500.000); mar1m (1:1.000.000)
mar2 - banco de dados batimétrico - Náutico
aero1 - banco aeronaútico cartográfico - Aviação
seus dados GIS - crie geodatabase separado para cada
escala cartográfica ex: aero500k (1:500.000);
eaero1m (1:1.000.000)
aero2 - banco aeroportos - Aviação
lr - banco referenciamento linear - Rodovias e estradas principais / Referenciamento Pipeline
ind - banco ambientes internos (Indoors)
Para uma featureclass muito grande, crie uma palavra-chave dbtune separada que aponte para tablespaces e arquivos de dados separados no armazenamento do banco, isso reduz a contenção no disco e melhora o I/O.
UN - Utility Network (Rede Utilitária)
- Cada Utility Network (UN) deve ser mantida em seu próprio geodatabase separado.
- Isso mantém o banco funcionando suavemente ao reduzir a contenção e
melhora o desempenho.
- Também facilita a gestão das tarefas de manutenção do banco,
como reconstruir índices e coletar estatísticas.
- Podemos facilmente clonar esse banco UN específico, usando backups do banco,
quando uma cópia é necessária para testes.
- Além disso, facilita mover o banco UN para um servidor diferente
- If something goes wrong and data is lost because of user mistakes,
you can recover just that specific UN database to a certain point in time.
- This is especially useful if the lost data can’t be recovered through
geodatabase versions.
- This setup helps meet important goals for data recovery,
called RPO (Recovery Point Objective) and RTO (Recovery Time Objective),
ensuring quick recovery and minimal data loss.
- For example, imagine there are two separate UN geodatabases:
one for gas, and one for water.
- If someone deletes data from the water database in a geodatabase
child version and then reconciles and posts the edits to the geodatabase
sde.default version, you can’t get that deleted data back from the sde.default
version anymore.
- But, if you realize the mistake soon enough, the database administrator can
restore the water database to a previous point in time.
- Doing this doesn’t affect the gas database edits at all.
- If instead, all UNs were stored in one geodatabase, deleting or losing data
from one UN could risk losing edits from all datasets in that geodatabase.
An Esri ® Technical Reference February• 2026
________________________________________________________________________________________________
FAQ - Frequent Asked Questions
🎯 How Load Large Featureclass Oracle Geodatabase
🎯 How Load Large Featureclass SQL Server Geodatabase
🎯 How to Move and Upgrade a WMX Repository Enterprise GDB
🎯 How to Move the WMX table JTX_JOB_ATTACHMENTS
🎯 Pro Branch Version & Oracle Table Compression
🎯 Oracle eGDB SDE Repo Upgrade Using Oracle Restore Point | Download PDF Version
🎯 ArcSDE User-Schema Geodatabases in Oracle are deprecated
🎯 How to Move the Oracle Enterprise Geodatabase with the Oracle Data Pump Utility
🎯 How to Move the PostgreSQL Enterprise Geodatabase with pg_dump and pg_restore
🎯 How to Move the SQL Server Enterprise Geodatabase with a database backup
🎯 How to Install Database Clients for ArcGIS
🎯 How to Install the Oracle Database Client for ArcGIS?
🎯 How to Install the SQL Server Client for ArcGIS?
🎯 How to Upgrade the PostgreSQL and PostGIS version for the Ent. Geodatabase on Windows
🎯 How to Upgrade the PostgreSQL and PostGIS version for the Ent. Geodatabase on Linux
🎯 How to Configure Windows Authentication for the PostgreSQL Enterprise Geodatabase
🎯 Linux Docker Installation - Esri Community
🎯 Linux Docker: Enterprise Geodatabase in PostgreSQL
🎯 Linux Docker: Enterprise Geodatabase in SQL Server
🎯 Linux Docker: Enterprise Geodatabase in Oracle
________________________________________________________________________________________________
Guide Books ArcGIS Pro
The guide books are intended for database administrators to help establish the product workspaces in an Enterprise Geodatabase.
ArcGIS Pro 3.x
🌎 Topographic Mapping
Best Practices Production Mapping 3.x Workspace in Oracle® October/21/2025
Best Practices Production Mapping 3.x Workspace in SQL Server® October/21/2025
Best Practices Production Mapping 3.x Workspace in PostgreSQL® October/21/2025
🔍 Data Reviewer (Classic) - Deprecated! Final Supported Version: ArcGIS Pro 3.6
Best Practices Data Reviewer 3.x Workspace in Oracle® October/21/2025
Best Practices Data Reviewer 3.x Workspace in SQL Server® October/21/2025
Best Practices Data Reviewer 3.x Workspace in PostgreSQL® October/21/2025
ArcGIS Data Reviewer FunctionalityDeprecation of Functionality in ArcGIS Data Reviewer
Final Supported Version: ArcGIS Pro 3.6 (released October 2025) is the last version
to support legacy tools like Reviewer Batch Jobs, Reviewer Sessions, and
the Reviewer Table.
Removal Version: Starting with ArcGIS Pro 3.7 (slated for April 2026),
these classic components will be completely removed and no longer available.
Reason for Change:
This aligns with the final retirement of ArcMap 10.8.x in March 2026,
as the Classic workspace was primarily designed for interoperability
with ArcMap.
Modern Alternative: Users must migrate to Attribute Rules-based workflows,
which utilize the Error Inspector for error management instead of the legacy
Reviewer Table.
📝 Workflow Manager (Classic)
- Deprecated! Final Supported Version: ArcGIS Pro 3.4
Best Practices Workflow Manager 3.x Workspace in Oracle® October/21/2025
Best Practices Workflow Manager 3.x Workspace in SQL Server® October/21/2025
Best Practices Workflow Manager 3.x Workspace in PostgreSQL® October/21/2025
ArcGIS Pro 3.4 is the final release that supports Workflow Manager (Classic).
🌍 Aviation - Charting / Airports
Best Practices Aviation 3.x Workspace in Oracle® October/21/2025
Best Practices Aviation 3.x Workspace in SQL Server® October/21/2025
Best Practices Aviation 3.x Workspace in PostgreSQL® October/21/2025
🌐 Maritime - Charting
Best Practices Maritime Charting 3.x Workspace in Oracle® October/21/2025
Best Practices Maritime Charting 3.x Workspace in SQL Server® October/21/2025
Best Practices Maritime Charting 3.x Workspace in PostgreSQL® October/21/2025
🌐 Maritime - Bathymetry
Best Practices Bathymetry 3.x Workspace in Oracle® October/21/2025
Best Practices Bathymetry 3.x Workspace in SQL Server® October/21/2025
Best Practices Bathymetry 3.x Workspace in PostgreSQL® October/21/2025
🌏 Linear Referencing - Roads & Highways / Pipeline Referencing
Best Practices LR 3.x Workspace in Oracle® October/21/2025
Best Practices LR 3.x Workspace in SQL Server® October/21/2025
Best Practices LR 3.x Workspace in PostgreSQL® October/21/2025
🌐 Indoors
Best Practices Indoors 3.x Workspace in Oracle® October/21/2025
Best Practices Indoors 3.x Workspace in SQL Server® October/21/2025
Melhores Práticas Indoors 3.x Workspace no PostgreSQL4 Outubro/21/2025
Lançamentos Anteriores ArcGIS Pro
________________________________________________________________________________________________
Guias ArcMap 10.8.x
Ciclo de Vida do ArcMap - Descontinuação 01 de Março de 2026
Os guias são destinados a administradores de banco de dados para ajudar a estabelecer os workspaces do produto em um Enterprise Geodatabase.
Nota: Clientes de Roads and Highways, Pipeline Referencing, GeoEvent, Indoors, Aviation podem seguir as mesmas melhores práticas nos Guias de Produção Mapping
ArcMap 10.8 20 de Fev, 2020
🌎 Production Mapping
Production Database
Melhores Práticas Production Mapping 10.8 Workspace no Oracle4
Melhores Práticas Production Mapping 10.8 Workspace no SQL Server40
Melhores Práticas Production Mapping 10.8 Workspace no PostgreSQL4
📕 Product Library
Melhores Práticas Product Library 10.8 Workspace no Oracle4
Melhores Práticas Product Library 10.8 Workspace no SQL Server4
Melhores Práticas Product Library 10.8 Workspace no PostgreSQL4
🔍 Data Reviewer
Melhores Práticas Data Reviewer 10.8 Workspace no Oracle4
Melhores Práticas Data Reviewer 10.8 Workspace no SQL Server4
Melhores Práticas Data Reviewer 10.8 Workspace no PostgreSQL4
📝 Workflow Manager
Melhores Práticas Workflow Manager 10.8 Workspace no Oracle4
Melhores Práticas Workflow Manager 10.8 Workspace no SQL Server4
Melhores Práticas Workflow Manager 10.8 Workspace no PostgreSQL4
🌐 Bathymetric Solution (BIS)
Melhores Práticas Bathymetry Solution (BIS) 10.8 Workspace no Oracle4
Melhores Práticas Bathymetry Solution (BIS) 10.8 Workspace no SQL Server4
🌐 Nautical Solution (NIS)
Melhores Práticas Nautical Solution (NIS) 10.8 no Oracle4
Melhores Práticas Nautical Product Library 10.8 no Oracle4
Melhores Práticas Nautical Solution (NIS) 10.8 no SQL Server4
Melhores Práticas Nautical Product Library 10.8 no SQL Server4
Lançamentos Anteriores ArcMap 10.7 / 10.6 / 10.5 / 10.4 / 10.3 / 10.2 / 10.1 / 10.0 .
________________________________________________________________________________________________
Lançamentos Anteriores - ArcGIS Pro 2.x - ArcMap 10.7 / 10.6 / 10.5 / 10.4 / 10.3 / 10.2 / 10.1 / 10.0 .
________________________________________________________________________________________________
Scripts Modelo para Bancos de Dados Enterprise MCS
Como instalar, configurar, fazer backup e manter Bancos de Dados Enterprise MCS Muito Grandes (VLMCSEDB), instruções detalhadas para carregamento de dados, backup e recuperação e para gerenciar múltiplos logins de usuários.
*Para Administradores Profissionais de Banco de Dados, requer experiência avançada em RDBMS e Geodatabase avançado.
Uma Referência Técnica Esri ® Fevereiro•2026
Geodatabases
Crie um geodatabase para cada workspace do produto de software.
pm - banco de dados de produção Production Mapping & Defense Mapping
- Mapeamento Topográfico - MGCP, TDS, etc.
seus dados gis - crie geodatabase separado para cada
escala cartográfica ex: pm50 (1:50k); pm250 (1:250k)
pl - workspace da product library - Product Library* (* Apenas ArcMap)
rev - workspace do data reviewer - Data Reviewer
wmx - workspace do workflow manager - Workflow Manager
mar1 - banco de dados cartográfico marítimo - Nautical
mar2 - banco de dados batimétrico - Nautical
aero1 - banco de dados cartográfico aeronáutico - Aviation
aero2 - banco de dados aeroportos - Aviation
lr - banco de dados linear referencing - Roads & Highways / Pipeline Referencing
ind - banco de dados indoors - Indoors
Para uma featureclass muito grande crie uma palavra-chave dbtune separada que aponte para tablespaces e datafiles separados no armazenamento do banco de dados, isso reduz a contenção do disco e melhora o I/O.
UN - Utility Network
- Cada Utility Network (UN) deve ser mantida em seu próprio geodatabase separado.
- Isso mantém o banco de dados funcionando suavemente ao reduzir a contenção e tornar
o desempenho melhor.
- Também facilita o gerenciamento das tarefas de manutenção do banco de dados, como reconstruir
índices e coletar estatísticas.
- Podemos facilmente clonar esse banco UN específico, usando backups do banco de dados,
quando uma cópia é necessária para testes.
- Além disso, facilita mover o banco UN para um servidor diferente se necessário.
ensuring quick recovery and minimal data loss.
- For example, imagine there are two separate UN geodatabases: one for gas,
and one for water.
- If someone deletes data from the water database in a geodatabase child
version and then reconciles and posts the edits to the geodatabase sde.default
version, you can’t get that deleted data back from the sde.default version
anymore.
- But, if you realize the mistake soon enough, the database administrator can
restore the water database to a previous point in time.
- Doing this doesn’t affect the gas database edits at all.
- If instead, all UNs were stored in one geodatabase, deleting or losing data
from one UN could risk losing edits from all datasets in that geodatabase.
An Esri ® Technical Reference October• 2025
Database Template Scripts
- the scripts cover the guide books best practices
- the scripts also have very advanced rdbms and advanced geodatabase best practices for VLMCSEDB that go beyond the guide books recommendations
Oracle Oracle 21c / 19c / 18c / 12c (12.2, 12.1) - For consolidation going forward adopt the Oracle 12c multi-tenant architecture: 1 CDB (container database): "mcs" hosts PDB's (pluggable databases): "pm", "pl", "rev", "wmx", "mar", "aero", "lr", "ind" each pluggable database is a standalone geodatabase with only the sde master repository installed - Discontinue the use of Esri user-schema geodatabases (1 oracle database - multiple geodatabases) ArcGIS Pro 3.x & ArcMap 10.9.x Oct 22, 2025 *New ArcGIS Pro 2.3 / 2.4 & ArcMap 10.7 / 10.8 Dec 18, 2020 ArcGIS Pro 2.2 & ArcMap 10.6 / 10.5 / 10.4 / 10.3 / 10.2 Aug 02, 2018 MCS Oracle 12c: Step-by-step Installation of Oracle RAC 12c on VMWare Nov 12, 2017 How to Load a Large Featureclass in an Oracle Geodatabase Oct 23, 2020 Oracle Geodatabase SDE Repository Upgrade using Oracle Restore Point Sep 06, 2022 How to Move eGDB using Oracle Data Pump Sep 13, 2022 Oracle 11gR2 ArcMap 10.2 / 10.1 March 27, 2014 - For consolidation, use Esri user-schema geodatabases (1 oracle database - multiple geodatabases)SQL Server SQL Server 2022 / 2019 / 2017 / 2016 (Windows/Linux) ArcGIS Pro 3.x & ArcMap 10.8.x June 21, 2022 *New ArcGIS Pro 2.3 / 2.4 & ArcMap 10.7 / 10.8 March 27, 2020 ArcGIS Pro 2.2 & ArcMap 10.6 / 10.5 / 10.4 / 10.3 / 10.2 / 10.1 November 13, 2017 MCS SQLServer: Step-by-step Installation of SQL Server 2016/2017 Cluster on a Windows Server 2016 Failover Cluster SQLServer Cluster on Windows Failover Cluster May 11, 2018 How to Load a Large Featureclass in a SQL Server Geodatabase Dec 03, 2020 PostgreSQL ArcGIS Pro 3.x & ArcMap 10.8.x Postgres on Windows July 05, 2023 *NewPostgres on Linux July 05, 2023 *New ArcGIS Pro 2.3 / 2.4 & ArcMap 10.7 / 10.8 Postgres on Windows March 20, 2020 Postgres on Linux March 20, 2020 ArcGIS Pro 2.2 & ArcMap 10.6 / 10.5 / 10.4 / 10.3 / 10.2 / 10.1 Postgres on Windows April 14, 2017 Postgres on Linux August 22, 2017
________________________________________________________________________________________________
Database Connections Best Practices
1. ArcSDE Application Server ("SDE Service") está obsoleto, use "direct connect". - Novidades no ArcMap 10.8 / 10.7 / 10.6 / 10.5 / 10.4 / 10.3 veja Databases & Geodatabases - Novidades no ArcGIS Pro - Plano de descontinuação para produtos ArcGIS 2. Nunca use a conexão do usuário "sde" em: Mapas (ex.: mxd's), ferramentas GP, Serviços de Mapa, Serviços de Feição, etc. 3. O usuário "sde" é apenas para administração da geodatabase (ex.: sde compress), o usuário sde é o proprietário do repositório da geodatabase 4. Nunca use conexões de usuário "data owner" em: Mapas (ex.: mxd's), ferramentas GP, Serviços de Mapa, Serviços de Feição, etc. - A menos que seja realmente necessário criar novas featureclasses ou tabelas na geodatabase - Mas é preciso entender as implicações - Para ferramentas de Geoprocessamento sempre prefira file geodatabases como áreas de trabalho temporárias, é mais rápido!!!! - O problema com múltiplas conexões de usuário "data owner" é que isso causa bloqueios de tabela no banco de dados - Isso pode levar a deadlocks e fazer com que as conexões do ArcGIS Desktop (ArcMap/Pro) e ArcGIS Server parem/congelem - Exceção é Branch Versioning, somente o data owner pode publicar o Feature Service 5. Como evitar o problema de "deadlock" então? - Se não estiver editando dados no ArcGIS, use uma conexão de usuário "viewer" para criar o mapa - Se estiver editando dados no ArcGIS, use uma conexão de usuário "editor" para criar o mapa - A conexão do usuário "viewer" tem permissão somente leitura (select) nas Featureclasses e Tabelas do usuário "data owner" - A conexão do usuário "editor" tem permissão leitura e escrita (select, insert, update e delete) nas Featureclasses e Tabelas do usuário "data owner" - Se estiver criando Serviços de Mapa do ArcGIS Server use uma conexão de usuário "viewer" para criar o mapa antes da publicação - Se estiver criando Serviços de Feição do ArcGIS Server use uma conexão de usuário "editor" para criar o mapa antes da publicação - Se estiver criando Serviços de Geoprocessamento do ArcGIS Server: se estiver editando dados use "editor", caso contrário use a conexão "viewer" 6. Explicação detalhada sobre "deadlock" - Oracle-Base deadlocks | Documentação Oracle deadlocks - Documentação SQLServer deadlocks - Documentação PostgreSQL deadlocks
________________________________________________________________________________________________
Links da Documentação do Banco de Dados
________________________________________________________________________________________________
Database Storage Best Practices
Why to use Multiple Tablespaces and RAID 10
• Large production enterprise geodatabase systems should employ a hardware striping solution. Your best disk and data organization strategies involve spreading your data across multiple disks.• With data spread across multiple disks, more spindles actively search for it. This can increase disk read time and decrease disk contention.• There are two main ways of achieving striping: tablespaces and redundant array of independent disks (RAID).• You can also combine the two by creating tablespaces within disk arrays.• You can employ data segregation strategies; keeping tables from indexes or certain types of tables from other tables will improve performance and alleviate administrative burdens.• Standard GIS storage recommendations favor keeping index and log files separate from vector and tabular business tables.• For performance reasons, it is better to position the business, feature, and spatial index tables separately and position tablespace data files based on their usage pattern.• For a multiversioned, highly active editing geodatabase, database files of the VERSIONS tablespace may be separated and dispersed across available disks to avoid I/O contention.• As a rule, you should create database files as large as possible based on the maximum amount of data you estimate the database will contain to accommodate future growth.• By creating large files, you can avoid file fragmentation and gain better database performance.• In many cases, you can allow data files to grow automatically; just be sure to limit autogrowth by specifying a maximum growth size that leaves some available hard disk space.• By putting different tablespaces on different disks, you can also help eliminate physical fragmentation of your files as they grow.
Famous white paper!!!!!• Fine-grained striping can be a bad choice for the tablespaces to which Oracle Server writes sort segments if a lot of applications processes will be sorting simultaneously. Fine-grained striping can a so yield surprisingly bad results for applications that use the Oracle's parallel query optimization (PQO) feature, because PQO's use of multiple query slaves manufactures its own high concurrency level even for a single user session.• If your database contains one or more small tablespaces for which large raw slices would be wasteful, you may wish to select one or two more standard raw slice sizes for those tablespaces. If you use more than one raw slice size, then make the sizes integral multiples and divisors of each other.• I/O performance—Each tablespace should contain segments whose I/O concurrency and I/O size characteristics are similar, to facilitate disk array size and stripe size selection. Grouping read-only segments into read-only tablespaces reduces backup and recovery data transfer volumes and reduces PCM lock maintenance.• Outage resilience—Small tablespaces allow offline tablespace maintenance with minimal application outage. The system tablespace cannot be taken off-line, and it cannot be dropped and recreated, so keeping as few segments there as possible minimizes the need for database downtime. Isolating rollback segments into as few tablespaces as possible reduces outage frequency, because any tablespace containing an on-line rollback segment cannot be taken off-line. Storing referentially related segments in small groups of tablespaces maximizes your ability to exploit the tablespace point-in-time recovery feature.• Space management—Isolating segments with short lifespans minimizes the impact of tablespace free space fragmentation that can block Oracle Server extent allocation. VLDB administrators profit from using tablespace default storage parameters instead of maintaining segment sizing parameters at the segment level.• Quota management—Oracle Server space quotas are administered to users by tablespace. Hence, you should assign groups of segments in one schema to a small group of tablespaces.• If you are using fine-grained striping for data files, then it is not necessary to separate indexes from data to distribute I/O load across disks. However, if your operational procedures include periodic index rebuilds, you may want to consider isolating indexes into their own tablespaces to minimize the possible impact of tablespace free space fragmentation caused by the drop index.
• This methodology assigns a specific load to various subsets of the database such as log files, rollback segments, index tablespaces, data tablespaces, etc. and assigns disks to these subsets based on their load profile.
________________________________________________________________________________________________
An Esri ® Technical Reference February• 2026
________________________________________________________________________________________________
