Replacing R Server with ML Server
This commit is contained in:
Jeroen ter Heerdt
Родитель
9f08ba53b0
Коммит
08a7b8e621
|
|
@ -1 +1,2 @@
|
|||
This solution has moved to [Microsoft/r-server-campaign-optimization](https://github.com/Microsoft/r-server-campaign-optimization)
|
||||
# Campaign Management
|
||||
This solution has moved to [Microsoft/r-server-campaign-optimization](https://github.com/Microsoft/r-server-campaign-optimization).
|
||||
|
|
@ -4,7 +4,7 @@ Script to trian and test thecustomer churn template with SQL + MRS
|
|||
|
||||
.DESCRIPTION
|
||||
This script will show the E2E work flow of customer churn machine learning
|
||||
templates with Microsoft SQL 2016 and Microsoft R services.
|
||||
templates with Microsoft SQL 2016 or later and Microsoft ML services.
|
||||
|
||||
Switch parameter ResetParmOnly allows you to reset the SQL database name.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,10 +1,10 @@
|
|||
This diretory contains R (Microsoft ML Server) codes for Energy Demand Forecasting Template with SQL Server 2016 R Services.
|
||||
This directory contains R (Microsoft ML Server) codes for Energy Demand Forecasting Template with SQL Server ML Services.
|
||||
The script main.R can be run from an R IDE, which will go through each step of the model development process of this template.
|
||||
The following scripts are included:
|
||||
|
||||
Script name|Description
|
||||
-----------|-----------
|
||||
dataPrepration.R|Fill missing data in the historical dataset
|
||||
dataPreparation.R|Fill missing data in the historical dataset
|
||||
featureEngineering.R|Compute features including month of year, hour of day, weekday/weekend, linear trend, Fourier components, lag, etc.
|
||||
trainModel.R|Train a Random Forest Regression model using the high performance analytics algorithm rxDForest in Microsoft ML Server (MRS)
|
||||
main.R|Step by step demonstration of the model development process
|
||||
|
|
|
|||
|
|
@ -1,7 +1,7 @@
|
|||
# Energy Demand Forecast Template with SQL Server ML Services
|
||||
Demand forecasting is an important problem in various domains including energy, retail, services, etc. Accurate demand forecasting helps companies conduct better production planning, resource allocation, and make other important business decisions. In the energy sector, demand forecasting is critical for reducing energy storage cost and balancing supply and demand.
|
||||
This template demonstrates how to use [SQL Server ML Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/what-is-sql-server-machine-learning) to build an end-to-end, on-prem solution for electricity demand forecasting. For a cloud-based solution using Cortana Analytics Suite(CAS), please see [CAS Solution Template: Demand Forecasting for Energy](https://gallery.cortanaanalytics.com/SolutionTemplate/Demand-Forecasting-for-Energy-1).The solution template includes a real time data simulator, feature engineering, model retraining, forecasting, and visualization.
|
||||
In this template with SQL Server R Services, we show two versions of implementation:
|
||||
In this template with SQL Server ML Services, we show two versions of implementation:
|
||||
|
||||
- **Model Development with Microsoft ML Server in R IDE**. Run the code in R IDE (e.g., RStudio, R Tools for Visual Studio) with data in SQL Server, and execute the computation in SQL Server.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
# Energy Demand Forecast Template with SQL Server 2016 R Services
|
||||
# Energy Demand Forecast Template with SQL Server ML Services
|
||||
* **Introduction**
|
||||
* **System Requirements**
|
||||
* **Workflow Automation**
|
||||
|
|
@ -18,12 +18,12 @@ You need SQL Server 2016 RC1 or later to deploy this template. SQL Server 2016 R
|
|||
* Install SQL Server 2016 on you own computer or server. You can use both Windows and SQL Server Authentication in this case. Windows Authentication is recommended as no additional firewall configuration is needed.
|
||||
* Follow instructions [here](https://azure.microsoft.com/en-us/documentation/articles/virtual-machines-provision-sql-server/) to provision a SQL server virtual machine in Azure. You will need an Azure subscription to do this. You can only use SQL Server Authentication to access a remote server.
|
||||
|
||||
If you are using SQL Server 2016 RC2, follow the post installation instructions [here](https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install) to set up SQL Server R Services.
|
||||
Follow the installation instructions [here](https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install) to set up SQL Server ML Services.
|
||||
|
||||
* **Login requirements.**
|
||||
The login you use to access the SQL Server and database needs to have the following permissions. **NOTE**: If the login is a member of the **sysadmin** server role, it has met all the requirements. You can check this in the “Security” section of the server using SQL Server Management Studio or Visual Studio.
|
||||
* Permission to create database. The login needs to be a member of the **dbcreator** server role to create new database.
|
||||
* Permissions to read data and run R scripts in the database. If you followed the post-installation configuration steps when setting up R Services, your login should already have these permissions.
|
||||
* Permissions to read data and run R scripts in the database. If you followed the post-installation configuration steps when setting up ML Services, your login should already have these permissions.
|
||||
* Access to SQL Server Agent. SQL Server Agent is used to schedule jobs in this template. Members of the **SQLAgentUserRole**, **SQLAgentReaderRole**, and **SQLAgentOperatorRole** fixed database roles in msdb, and members of the **sysadmin** fixed server role have access to SQL Server Agent. A user that does not belong to any of these roles cannot use SQL Server Agent. For more information on the roles used by SQL Server Agent, see [Implement SQL Server Agent Security](https://docs.microsoft.com/en-us/sql/ssms/agent/implement-sql-server-agent-security).
|
||||
* **Local client requirements.**
|
||||
On the computer where you will run the deployment script, you need the following programs installed:
|
||||
|
|
|
|||
|
|
@ -1 +1,2 @@
|
|||
# Fraud detection
|
||||
This solution has moved to [Microsoft/r-server-fraud-detection](https://github.com/Microsoft/r-server-fraud-detection).
|
||||
|
|
@ -1,6 +1,6 @@
|
|||
# Galaxies classification with Deep Learning from Mirosoft ML using SQL Server R Services
|
||||
# Galaxies classification with Deep Learning from Microsoft ML using SQL Server ML Services
|
||||
|
||||
This sample provides the supporting SQL and R scripts for the blogpost [How six lines of code + SQL Server can bring Deep Learning to ANY App](https://blogs.technet.microsoft.com/dataplatforminsider/2017/01/05/how-six-lines-of-code-sql-server-can-bring-deep-learning-to-any-app/).
|
||||
This sample provides the supporting SQL and R scripts for the blog post [How six lines of code + SQL Server can bring Deep Learning to ANY App](https://blogs.technet.microsoft.com/dataplatforminsider/2017/01/05/how-six-lines-of-code-sql-server-can-bring-deep-learning-to-any-app/).
|
||||
|
||||
**Data**: [Galaxy Zoo](https://www.galaxyzoo.org/) project was used as source of labeled training data.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,10 +1,10 @@
|
|||
# SQL Server R Services - Performance Test Samples
|
||||
# SQL Server ML Services - Performance Test Samples
|
||||
|
||||
The sample runtest.R script provided in this repository demonstrates the benefits of following various tips for improving performance of R script running in SQL compute context.
|
||||
|
||||
The test script helps to evaluate the performance gains that can be reailized when following the various tips outlined in the <a href="https://docs.microsoft.com/en-us/sql/advanced-analytics/r/sql-server-r-services-performance-tuning">SQL Server ML Services Performance Tuning document</a>.
|
||||
|
||||
To run these tests, the user needs access to SQL Server 2016 with R services enabled. The user is also assumed to be familiar with some basic concepts of using SQL Server R Services.
|
||||
To run these tests, the user needs access to SQL Server with ML Services enabled. The user is also assumed to be familiar with some basic concepts of using SQL Server ML Services.
|
||||
|
||||
The directory contents are as follows:
|
||||
|
||||
|
|
@ -41,7 +41,7 @@ To create the database, download the 2 files from the links below into some fold
|
|||
> https://sqlrperftuning.blob.core.windows.net/perftuningdb/PerfTuning2.bak
|
||||
|
||||
**Steps**
|
||||
> - Ensure that you have access to SQL Server 2016 with R Service enabled. You should also have permission to create or access the PerfTuning database.
|
||||
> - Ensure that you have access to SQL Server with ML Services enabled. You should also have permission to create or access the PerfTuning database.
|
||||
> - Create the PerfTuning database and add tables using the attached R and SQL scripts. You can either run createall.cmd or run the other db and table creating scripts one by one. To expedite, you can skip running the scripts by downloading the database using links above and restoring.
|
||||
> - Install the dependant RODBC package, if not installed. Ensure that it is installed in the right library where Microsoft RevoScaleR package was installed. (If you ran creatall.cmd, it will install this package. Update the file if lib path needs to be specified)
|
||||
> - Update the runtests.R file to match your connection string and data directories.
|
||||
|
|
|
|||
|
|
@ -1,4 +1,5 @@
|
|||
This is the R (Microsoft ML Server) code for Predictive Maintenance template using SQL Server R Services.
|
||||
# Predictive Maintenance
|
||||
This is the R (Microsoft ML Server) code for Predictive Maintenance template using SQL Server ML Services.
|
||||
|
||||
The template solves the following problems:
|
||||
|
||||
|
|
|
|||
|
|
@ -34,7 +34,7 @@ The solutions are demonstrated using an aircraft engine data, with the following
|
|||
|
||||
For a full description of the template, please refer to the [template](https://gallery.cortanaanalytics.com/Collection/Predictive-Maintenance-Template-3) in Cortana Analytics gallery.
|
||||
|
||||
In this template with SQL Server R Services, we show two version of implementation:
|
||||
In this template with SQL Server ML Services, we show two version of implementation:
|
||||
|
||||
- **Model Development with Microsoft ML Server in R IDE**. Run the code in R IDE (e.g., RStudio, R Tools for Visual Studio) with data in SQL Server, and execute the computation in SQL Server.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
#Predictive Maintenance Template with SQL Server 2016 R Services
|
||||
#Predictive Maintenance Template with SQL Server ML Services
|
||||
------------------
|
||||
|
||||
* **Introduction**
|
||||
|
|
@ -41,11 +41,11 @@ The SQL procedures can be executed in SQL Server environment (such as **SQL Serv
|
|||
|
||||
To run the scripts, you must prepare the following environment:
|
||||
|
||||
* An instance of SQL Server 2016 (Enterprise or Developer edition) CTP 3 or later, with SQL Server R Services installed and configured
|
||||
* An instance of SQL Server 2016 (Enterprise or Developer edition) CTP 3 or later, with SQL Server ML Services installed and configured
|
||||
* A SQL login and password. The SQL login must have permissions to execute R scripts
|
||||
* A database on the instance in which the login has been granted the permission to create and execute stored procedures
|
||||
* The "plyr" and "zoo" R packages are needed. To install packages into the R Services environment please read here: https://docs.microsoft.com/en-us/sql/advanced-analytics/r/install-additional-r-packages-on-sql-server
|
||||
* For more information about SQL Server 2016 and SQL Server R Services, please visit: https://docs.microsoft.com/en-us/sql/advanced-analytics/what-s-new-in-sql-server-machine-learning-services
|
||||
* The "plyr" and "zoo" R packages are needed. To install packages into the ML Services environment please read here: https://docs.microsoft.com/en-us/sql/advanced-analytics/r/install-additional-r-packages-on-sql-server
|
||||
* For more information about SQL Server ML Services, please visit: https://docs.microsoft.com/en-us/sql/advanced-analytics/what-s-new-in-sql-server-machine-learning-services
|
||||
|
||||
###WORKFLOW AUTOMATION
|
||||
-------------------
|
||||
|
|
|
|||
|
|
@ -4,7 +4,7 @@ Script to trian and test the preventive maintenance template with SQL + MRS
|
|||
|
||||
.DESCRIPTION
|
||||
This script will show the E2E work flow of Preventive Maintenance machine learning
|
||||
templates with Microsoft SQL 2016 and Microsoft R services.
|
||||
templates with Microsoft SQL and ML services.
|
||||
|
||||
Switch parameter Score allows you to score the production data with seleted model
|
||||
Switch parameter ResetParmOnly allows you to reset the SQL related credentials
|
||||
|
|
|
|||
|
|
@ -1,5 +1,5 @@
|
|||
#SQL R Services: Predictive Maintenance Modeling Guide
|
||||
In this template, we demonstrate how to develop a Predictive Maintenance solution with SQL Server 2016 R Services where the process is aligned with the existing R Notebook published in the [Cortana Intelligence Gallery](https://gallery.cortanaintelligence.com/Notebook/Predictive-Maintenance-Modelling-Guide-R-Notebook-1) but works with a larger dataset.
|
||||
# SQL ML Services: Predictive Maintenance Modeling Guide
|
||||
In this template, we demonstrate how to develop a Predictive Maintenance solution with SQL Server ML Services where the process is aligned with the existing R Notebook published in the [Cortana Intelligence Gallery](https://gallery.cortanaintelligence.com/Notebook/Predictive-Maintenance-Modelling-Guide-R-Notebook-1) but works with a larger dataset.
|
||||
|
||||
##In this template:
|
||||
* There are 5 data sources namely: telemetry, errors, maintenance, machines, failures
|
||||
|
|
@ -9,7 +9,7 @@ In this template, we demonstrate how to develop a Predictive Maintenance solutio
|
|||
* Data visualization and multi-class classification model is done via R code running on SQL Server
|
||||
|
||||
##Implementation prerequisites:
|
||||
* SQL Server 2016 with R Services: https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install
|
||||
* SQL Server with ML Services: https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install
|
||||
|
||||
* R IDE such as R Studio or R Tools for Visual Studio to access the data from the server: https://docs.microsoft.com/en-us/machine-learning-server/r-client/what-is-microsoft-r-client
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
#Retail Customer Cross-sell Template with SQL Server Machine Learning (R) Services
|
||||
# Retail Customer Cross-sell Template with SQL Server Machine Learning ML Services
|
||||
|
||||
In this template, we demonstrate how to develop and deploy end-to-end customer cross-sell prediction models with SQL Server Machine Learning Services.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
This is the R (Microsoft ML Server) code for Retail Forecasting template using SQL Server R Services.
|
||||
This is the R (Microsoft ML Server) code for Retail Forecasting template using SQL Server ML Services.
|
||||
|
||||
Two modeling solutions are provided for this template :
|
||||
|
||||
|
|
|
|||
|
|
@ -1,8 +1,8 @@
|
|||
#Retail Forecasting Template with SQL Server 2016 R Services
|
||||
# Retail Forecasting Template with SQL Server ML Services
|
||||
|
||||
In this template, we demonstrate how to develop and deploy end-to-end Retail Forecasting solutions with [SQL Server 2016 R Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/what-is-sql-server-machine-learning).
|
||||
In this template, we demonstrate how to develop and deploy end-to-end Retail Forecasting solutions with [SQL Server ML Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/what-is-sql-server-machine-learning).
|
||||
|
||||
Accurate and timely forecast in retail business drives success. It is an essential enabler of supply and inventory planning, product pricing, promotion, and placement. This template will demonstrate how to build a retail forecasting solution with SQL and Microsoft R services using the sales data from the retail industry. The sample data has been anonymized and transformed before being used in this sample.
|
||||
Accurate and timely forecast in retail business drives success. It is an essential enabler of supply and inventory planning, product pricing, promotion, and placement. This template will demonstrate how to build a retail forecasting solution with SQL and Microsoft ML services using the sales data from the retail industry. The sample data has been anonymized and transformed before being used in this sample.
|
||||
|
||||
The input data schema is as following:
|
||||
|
||||
|
|
@ -10,7 +10,7 @@ The input data schema is as following:
|
|||
|
||||
For a full description of the template, please refer to the [template](https://gallery.cortanaintelligence.com/Experiment/Retail-Forecasting-Step-1-of-6-data-preprocessing-5) in Cortana Analytics gallery.
|
||||
|
||||
In this template with SQL Server R Services, we show two version of implementation:
|
||||
In this template with SQL Server ML Services, we show two version of implementation:
|
||||
|
||||
- **Model Development with Microsoft ML Server in R IDE**. Run the code in R IDE (e.g., RStudio, R Tools for Visual Studio) with data in SQL Server, and execute the computation in SQL Server.
|
||||
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
#Retail Forecasting Template with SQL Server 2016 R Services
|
||||
#Retail Forecasting Template with SQL Server ML Services
|
||||
------------------
|
||||
|
||||
* **Introduction**
|
||||
|
|
@ -14,7 +14,7 @@
|
|||
##INTRODUCTION
|
||||
------------
|
||||
|
||||
This template demonstrates how to build a pipeline that automatically provides weekly retail forecasts of the next 52 weeks for each store and each product using [SQL Server R Services.](https://docs.microsoft.com/en-us/sql/advanced-analytics/what-is-sql-server-machine-learning).
|
||||
This template demonstrates how to build a pipeline that automatically provides weekly retail forecasts of the next 52 weeks for each store and each product using [SQL Server ML Services.](https://docs.microsoft.com/en-us/sql/advanced-analytics/what-is-sql-server-machine-learning).
|
||||
|
||||
Two modeling solutions are provided for this template :
|
||||
|
||||
|
|
@ -37,11 +37,11 @@ The SQL procedures can be executed in SQL Server environment (such as **SQL Serv
|
|||
|
||||
To run the scripts, you must prepare the following environment:
|
||||
|
||||
* An instance of SQL Server 2016 CTP 3 or later, with SQL Server R Services installed and configured
|
||||
* An instance of SQL Server 2016 CTP 3 or later, with SQL Server ML Services installed and configured
|
||||
* A SQL login and password. The SQL login must have permissions to execute R scripts
|
||||
* A database on the instance in which the login has been granted the permission to create and execute stored procedures
|
||||
|
||||
* For more information about SQL Server 2016 and SQL Server R Services, please visit:
|
||||
* For more information about SQL Server ML Services, please visit:
|
||||
https://docs.microsoft.com/en-us/sql/advanced-analytics/what-s-new-in-sql-server-machine-learning-services
|
||||
|
||||
###WORKFLOW AUTOMATION
|
||||
|
|
|
|||
|
|
@ -4,7 +4,7 @@ Script to trian, test and evaluate the retail forecasting template with SQL + MR
|
|||
|
||||
.DESCRIPTION
|
||||
This script will show the E2E work flow of retail forecasting machine learning
|
||||
templates with Microsoft SQL 2016 and Microsoft R services.
|
||||
templates with Microsoft SQL ML services.
|
||||
|
||||
For the detailed description, please read README.md.
|
||||
#>
|
||||
|
|
|
|||
|
|
@ -1,8 +1,8 @@
|
|||
# SQL Optimization Tips and Tricks for Analytics Services
|
||||
# Resume matching: SQL Optimization Tips and Tricks for Analytics Services
|
||||
|
||||
## Introduction
|
||||
|
||||
In SQL Server 2016, a new function, which is called R services, has been added. SQL Server 2016 R Services provides a platform for operationalize R scripts using T-SQL to develop and deploy intelligent applications. This Markdown file will describe the design and key optimization techniques for a resume-matching scenario that demonstrates how we can find the best candidates for a job opening among millions of resumes within a few seconds.
|
||||
In SQL Server 2016, a new function, which what was called R services, has been added. Later it was renamed to ML Services. SQL Server ML Services provides a platform for operationalize R and Python scripts using T-SQL to develop and deploy intelligent applications. This Markdown file will describe the design and key optimization techniques for a resume-matching scenario that demonstrates how we can find the best candidates for a job opening among millions of resumes within a few seconds.
|
||||
|
||||
## Use Case
|
||||
|
||||
|
|
@ -38,9 +38,9 @@ Three synthetic datasets were generated for the purpose of demonstrating the sce
|
|||
|
||||
## Enable to run R script within SQL query
|
||||
|
||||
In order to run R script within SQL query, we first need to enable this functionality. You can enable the R Services when you are provisioning the VM. If you forgot to do so, we are showing the detailed steps to enable SQL Server to run R code.
|
||||
In order to run R script within SQL query, we first need to enable this functionality. You can enable the ML Services when you are provisioning the VM. If you forgot to do so, we are showing the detailed steps to enable SQL Server to run R code.
|
||||
|
||||
* **Step 1:** Run the following SQL query to explicitly enable the R Services feature on SQL Server; otherwise, it will not be possible to invoke R scripts even if the feature has been installed by setup.
|
||||
* **Step 1:** Run the following SQL query to explicitly enable the ML Services feature on SQL Server; otherwise, it will not be possible to invoke R scripts even if the feature has been installed by setup.
|
||||
```sql
|
||||
Exec sp_configure 'external scripts enabled', 1
|
||||
Reconfigure with override
|
||||
|
|
@ -81,7 +81,7 @@ During setup of the SQL Server, 20 new Windows user accounts are created for the
|
|||
|
||||
## Implementation
|
||||
|
||||
In this section, we will describe in great detail of the implementation on a SQL Server 2016 with R Services to handle the resume matching problem. The implementation includes the optimizations that have been applied on this specific machine, R code to train a matching model, to use the matching model, and a PowerShell script to launch multiple batch scoring concurrently. All those components are organized in the format of a few SQL scripts. Those scripts are used to configure the SQL server, optimize the server for this data science scenario, train the prediction model and score for each project (job). All those scripts are placed under the "***SQLR***" folder.
|
||||
In this section, we will describe in great detail of the implementation on a SQL Server ML Services to handle the resume matching problem. The implementation includes the optimizations that have been applied on this specific machine, R code to train a matching model, to use the matching model, and a PowerShell script to launch multiple batch scoring concurrently. All those components are organized in the format of a few SQL scripts. Those scripts are used to configure the SQL server, optimize the server for this data science scenario, train the prediction model and score for each project (job). All those scripts are placed under the "***SQLR***" folder.
|
||||
|
||||
We will describe the implementation step by step and show the SQL queries as well.
|
||||
|
||||
|
|
@ -705,7 +705,7 @@ A prediction model named "_rxBTrees_" will be saved in table "_dbo.Classificatio
|
|||
|
||||
The trained model can now be used to find the best candidates given a now job. The input data for the model will be a project (job) ID, and the model will generate the features for all resume-job pairs and use the trained prediction model for scoring.
|
||||
|
||||
One of the primary benefits of SQL Server is its ability to handle a very large volume of rows in parallel. We first split the matching into a few tasks and each workload group will process one task. Furthermore, SQL Server with R Services can query the database within the R code to perform selection, joining and aggregations in parallel as well. We also created a saved procedure for prediction. The detail of the query please refer to file "***step5_score_for_matching.sql***".
|
||||
One of the primary benefits of SQL Server is its ability to handle a very large volume of rows in parallel. We first split the matching into a few tasks and each workload group will process one task. Furthermore, SQL Server with ML Services can query the database within the R code to perform selection, joining and aggregations in parallel as well. We also created a saved procedure for prediction. The detail of the query please refer to file "***step5_score_for_matching.sql***".
|
||||
|
||||
```sql
|
||||
USE ResumeMatching
|
||||
|
|
@ -890,19 +890,19 @@ The first select query will show you the total duration for scoring 1.1 million
|
|||
|
||||
## Conclusion
|
||||
|
||||
SQL Server 2016 with R Service provides a scalable solution to handle the resume matching use case. In this tutorial, we used in-memory table, soft-NUMA, resource pool, and resource governance techniques to optimize the computation on SQL server. By applying those optimization techniques, we have achieved to score **1.1 million** rows of data (with 100 features) within **8.5 seconds** on a 20 cores machine.
|
||||
SQL Server ML Services provides a scalable solution to handle the resume matching use case. In this tutorial, we used in-memory table, soft-NUMA, resource pool, and resource governance techniques to optimize the computation on SQL server. By applying those optimization techniques, we have achieved to score **1.1 million** rows of data (with 100 features) within **8.5 seconds** on a 20 cores machine.
|
||||
|
||||
## Useful References
|
||||
|
||||
[Configure and Manage Advanced Analytics Extensions](https://docs.microsoft.com/en-us/sql/advanced-analytics/r/managing-and-monitoring-r-solutions)
|
||||
|
||||
[Set up SQL Server R Services (In-Database)](https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install)
|
||||
[Set up SQL Server ML Services (In-Database)](https://docs.microsoft.com/en-us/sql/advanced-analytics/install/sql-r-services-windows-install)
|
||||
|
||||
[Use sqlBindR.exe to Upgrade an Instance of R Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/r/use-sqlbindr-exe-to-upgrade-an-instance-of-sql-server)
|
||||
[Use sqlBindR.exe to Upgrade an Instance of ML Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/r/use-sqlbindr-exe-to-upgrade-an-instance-of-sql-server)
|
||||
|
||||
[Recommendations and guidelines for the "max degree of parallelism" configuration option in SQL Server](https://support.microsoft.com/en-us/help/2806535/recommendations-and-guidelines-for-the-max-degree-of-parallelism-confi)
|
||||
|
||||
[Resource Governance for R Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/administration/resource-governance)
|
||||
[Resource Governance for ML Services](https://docs.microsoft.com/en-us/sql/advanced-analytics/administration/resource-governance)
|
||||
|
||||
[How to: Configure SQL Server to Use Soft-NUMA](https://docs.microsoft.com/en-us/previous-versions/sql/sql-server-2008-r2/ms345357(v=sql.105))
|
||||
|
||||
|
|
|
|||
Загрузка…
Ссылка в новой задаче