Retail:Retail Performance Profile
Contents |
 | This is a work in progress document |
Introduction
This page gives an overview of supported customer volumes by the OB Commerce solution. The information provided here can be used to validate if a specific business case is covered by our current solutions.
The numbers provided in the page here are based on current customers using the OB Commerce solutions or have been tested using combination of unit tests, multi-user/client/server stress tests using jmeter and selenium using one application server and one separate database server. For the high volume performance tests we have done minor tuning of the system and database. See the Installation wiki page for detailed information on installing Openbravo, and see the Performance wiki page for information on tuning Openbravo's technology stack.
The reference volumes are described from different perspectives:
- Store configuration: the number of POS systems by store, number of stores
- Masterdata: the volume of master data records and also the mass change volume supported
- Transaction volume: number of new tickets, number of ticket lines per ticket
An important aspect with higher volume customer cases is the remote data concept. With remote data the WebPOS client will not load product and customer data in the local WebPOS client but will access/query them remotely.
When deploying Openbravo in a higher volume environment it can make sense to investigate a multi-server deployment. A multi-server architecture makes it possible to spread the load and increase robustness of the overall solution.
Openbravo Commerce in larger Enterprise solutions
The Openbravo Commerce release RR15Q4 deliver important architectural improvements which make Openbravo Commerce very well suited for large enterprises. Areas which have been improved considerably are: transactional performance, allowing multi-server distributed architectures and supporting remote data handling.
Customer Characteristics
This document provides reference numbers for three customer profiles:
- Small: company with a few stores (1-3) without central HQ, HQ function is possibly executed in the store. The company uses the complete Openbravo functional stack including warehousing and finance.
- Medium: company with 10-30 stores, there is a central HQ function. The company uses the complete Openbravo functional stack including warehousing and finance.
- Large: company with multiple 100s of stores, there is a central HQ function. Openbravo is mainly used for the front office function, possibly including warehousing. Finance functionality is often handled through a dedicated non-Openbravo finance application.
Server Architecture
For the different customer profiles we advice different types of system architectures:
- Small: host the database and application on one system hosted in the cloud
- Medium: two server solution: database and application on different servers, hosted in the cloud
- Large: multi-server environment, larger stores have a dedicated store server to balance the load for login and data querying. Smaller stores can share one store server. One central transaction server (load balanced) to receive transactions. Replication between databases. Store servers are single system with database and application on one system, the transactional server has application and database separated on different servers, the central transaction server can be load balanced for the application.
Customer Reference Profile
Store Reference Sizing
The store reference sizing discusses the reference volumes related to number of stores, the number of pos systems etc.
| Performance Indicator | Small Enterprise | Medium Enterprise | Larger Enterprise |
|---|---|---|---|
| Total #Stores | <5 | 5-75 | 75-300 |
| Total #POS | 5 | 150 | 3000 |
| Average #POS/Store | 1 | 5 | 30 |
| Maximum #POS/Store | 1 | 10 | 750 |
Transaction Reference Volume
The transactional reference volumes gives an overview of the transactional volume we can handle with our current architecture.
| Performance Indicator | Small Enterprise | Medium Enterprise | Larger Enterprise |
|---|---|---|---|
| Total #tickets in history | 10000 | 2.5 million | 75 million |
| Average #tickets/day | 50 | 5000 | 100000 |
| Average #lines/ticket | 2 | 3 | 3 |
| Maximum #tickets/minute | 5 | 25 | 1750 |
| Maximum #tickets/day | 500 | 15000 | 350000 |
Master Data Reference Volume
Master data is data related to products, discounts, business partners etc. Master data can be changed and updated frequently centrally or through the WebPOS clients. Master data changes are distributed to WebPOS clients and store servers automatically.
| Performance Indicator | Small Enterprise | Medium Enterprise | Larger Enterprise |
|---|---|---|---|
| #Customers | <500 | 10000-30000 | 5-10 million |
| #Products | 50 | 500-1000 | 100000 upto 10 million |
| #Discounts | ? | ? | ? |
| #Product Prices | ? | ? | ? |
| #Master Data Changes avg/day | ? | ? | ? |
| #Master Data Changes max/day | ? | ? | ? |
Openbravo Commerce Modules for higher volume environments
| | Todo: add module version numbers |
In the 15Q4 release we extended our testing of extension modules as part of our continuous build environment which tests modules and functionality for every change we do. Many modules are now being automatically tested for their valid operation with remote master data enabled. Making them suitable for usage in an environment with large volumes of master data.
In the 2015 4th quarter we will further extend our continuous build environment to include automatic testing of Openbravo in high volume transaction and master data environment. Moving manual tests we have been executing to an automatic test environment.
- Mobile Physical Inventory
- Complementary Products
- Copy Retail Store
- Stock Criteria
- Retail Sessions
- Stock Validation
- Multi Tax Category
- Multi-UPC
- Mobile Procurement
- Product Description
- Tax Exempt
- Mobile Warehouse
The following modules have been tested/deployed in medium sized enterprises but have not yet been tested for the volumes used in large enterprises: