jsforce_downloader

jsforce_downloader

Extract report data from Salesforce into a comma separated file. This package includes 4 components that can be used independantly.

• nodejs library to download Salesforce reports that have a date filter.
• jsforce_downloader - command line utility to download Salesforce reports. (command line wrapper of the downloader).
• jsforce_s3_downloader - command line utility to download Salesforce reports directly to S3.
• jsforce_downloader_metadata - command line utility to display the metadata of a Salesforce report. Use this to inspect the fields and filters of a report. It also generates the SQL to create a table in MySQL to save this data.

Features

• Download more than 2000 details rows.
• Only extracts the detail rows (T!T) and ignores all group/summary sections.
• Exported as a CSV with the displayed value and the underlying value.
• Asynchronous reports are used to avoid the Salesforce limit on synchronous reports per hour.
• Parallel downloads to speed up the extract.
• Supports Tabular, Matrix and Summary report types.
• Support for AWS Lambda. Run this downloader in AWS Lambda. A Lambda event handler is provided.
• Support for AWS S3. Upload the downloaded data directly to S3 (no temp files needed on local machine).

Requirements

• The underlying object must have a date field that can be used as a standard date filter.
• Typically this can be the object's CreatedDate.

How to install

Install jsforce_downloader and jsforce_downloader_metadata.

npm install -g jsforce_downloader

Optionally - if you are using the AWS S3 feature, install the AWS SDK and set environment variables AWS_ACCESS_KEY, AWS_SECRET_KEY.

npm install -g aws-sdk

Environment variables to login

The library and utilities rely on the environment variables to store the username and password. If you are writing your own nodejs program, you can pass these during initialization.

SF_USER="[e-mail]"
SF_PASSWD_WITH_TOKEN="password";

The security token is required since the app does not support OAuth sign in. To get your security token, logon to Salesforce. At the top navigation bar go to your name > Setup > Personal Setup > My Personal Information > Reset My Security Token.

To use your token, if your password is mypassword, and your security token is XXXXXXXXXX, then set SF_PASSWD_WITH_TOKEN to "mypasswordXXXXXXXXXX" to log in. Your security token is reset and sent to your email whenever you change your password.

On the Mac OS X, I add the following lines to ~/.profile and restart Terminal.

export SF_USER="[e-mail]"
export SF_PASSWD_WITH_TOKEN="passwordTOKEN"

In Windows, you can follow the instructions to set environment variables. Restart your command or Powershell window after you set the environment variables.

If you are saving the output to S3 (OUTPUTTO="s3"), you should set the following environment variables. export AWS_ACCESS_KEY="access key id" export AWS_SECRET_KEY="secret for access key"

Command line tools: How to run jsforce_downloader_metadata

This will display all the columns and filters of a report. The metadata is saved as a JSON file. It will

jsforce_downloader_metadata {reportid}
jsforce_downloader_metadata 00OE0000002wlroMAA

This creates the file ReportOutput_00OE0000002wlroMAA.json. This file has the metadata for the report - so you can easily find the index of the column to display, {Report Section of the Fact Map} and report filters.

Command line tools: How to run jsforce_downloader

Preparation to download a report, you need

• The report ID (get this from the Salesforce URL when you open the report).
• The name of the date field - e.g. Case.CreatedDate to slice up the report into daily chunks. This does not have to be in the report.
• The zero-based index of column that is displayed while extracting (helps you keep track of the progress.) If you aren't sure, use 0.
• The section of the report that you want to see. This is explained in the Salesforce Analytics REST API guide - in the section decode the Fact Map. The pattern for the fact map keys varies by report format as shown in this table.

Report Fact map key pattern format

Tabular T!T: The grand total of a report. Both record data values and the grand total are represented by this key. Summary !T: T refers to the row grand total. Matrix !.

Each item in a row or column grouping is numbered starting with 0. Here are some examples of fact map keys:

 0!T The first item in the first-level grouping.
 1!T The second item in the first-level grouping.
 0_0!T The first item in the first-level grouping and the first item in the second-level grouping. 
 0_1!T The first item in the first-level grouping and the second item in the second-level grouping. 

  

To download a report, you need jsforce_downloader {reportid} {datefield} {index of field to display} {start date YYYY-MM-DD} {end date YYYY-MM-DD} [{MAX_Concurrent} [{Report section of the Fact Map}]]

Example:

      $ jsforce_downloader 00OE0000002wlroMAA Labor__c.CreatedDate 5 2016-01-01 2016-01-05 4 'T!T'

      Labor__c.CreatedDate 5 2016-01-01 2016-01-05
      Starting here....
      Report:00OE0000002wlroMAA
      Output to:ReportOutput_00OE0000002wlroMAA_20160101_to_20160105_20160413134312
      Start:2016-01-01
      End:2016-05-01
      Logged into Salesforce
      username: [e-mail](Sample User)
      0:Start Range: 2016-01-01T00:00:00-08:00 - 2016-01-01T23:59:59-08:00
      1:Start Range: 2016-01-02T00:00:00-08:00 - 2016-01-02T23:59:59-08:00
      2:Start Range: 2016-01-03T00:00:00-08:00 - 2016-01-03T23:59:59-08:00
      3:Start Range: 2016-01-04T00:00:00-08:00 - 2016-01-04T23:59:59-08:00
      4:Start Range: 2016-01-05T00:00:00-08:00 - 2016-01-05T23:59:59-08:00
      1:Returned Range: 2016-01-02T00:00:00-08:00 - 2016-01-02T23:59:59-08:00:Success
      84 records
      First: L-5156083 a0iE000000MiTNLIA3
      Last : L-5156837 a0iE000000MiUMMIA3
      Package size:83
      ....
      4:Returned Range: 2016-01-05T00:00:00-08:00 - 2016-01-05T23:59:59-08:00:Success
      665 records
      First: L-5172547 a0iE000000MihJHIAZ
      Last : L-5184790 a0iE000000Mir8jIAB
      Package size:664
      =============================
      Report:00OE0000002wlroMAA
      Date range:2016-01-01 to 2016-01-05
      Output to:ReportOutput_00OE0000002wlroMAA_20160101_to_20160105_20160413134312
      Done:1755 records written.
      Async reports requested:5 - (succeeded:5,failed:0).

This creates the file ReportOutput_00OE0000002wlroMAA_20160101_to_20160105_20160413134312.csv.

Command line tools: How to run jsforce_s3_downloader

To download a report, you need jsforce_s3_downloader {reportid} {datefield} {index of field to display} {start date YYYY-MM-DD} {end date YYYY-MM-DD} {s3 bucket} {s3 path} [{aws region}]

Example:

$ jsforce_s3_downloader 00OE0000002wlroMAA Labor__c.CreatedDate 5 2016-01-01 2016-01-04 monima test us-east-1
Switching AWS region to us-east-1
Starting here....
Report:00OE0000002wlroMAA
Output to:ReportOut_00OE0000002wlroMAA_20160101-20160104_20160418030436.csv
Start:2016-01-01
...
707 records
First row: (L-5158662,a0iE000000MiWcxIAF)
Last row : (L-5172382 a0iE000000MihGAIAZ)
=============================
Report       :00OE0000002wlroMAA
Date range   :2016-01-01 to 2016-01-04
Output to    :ReportOut_00OE0000002wlroMAA_20160101-20160104_20160418030436.csv
Done         :1087 records written.
Async reports:4 - (succeeded:4,failed:0).
Successfully uploaded data to monima/monima/ReportOut_00OE0000002wlroMAA_20160101-20160104_20160418030436.csv

Using the library in your NodeJS program.

Configuration of the library

var config = {
    MAX_CONCURRENT: 30, 
    // 30 parallel async report requests
    
    WAIT_BETWEEN_REQUESTS: 1000, 
    // 1000 milliseconds
    
    REPORTSECTION: "T!T", 
    // REPORTSECTION - The section of the report that you want to see. This is explained in the 
    // [Salesforce Analytics REST API guide](https://resources.docs.salesforce.com/sfdc/pdf/salesforce_analytics_rest_api.pdf) 
    // - in the section decode the Fact Map. 
    
    WRITE_TEMP_FILES: !fs.existsSync('./tmp'), 
    // Store output of each async report to the tmp subdirectory.
    
    SFOptions: {
        loginUrl: "https://login.salesforce.com"
    }, 
    // Initialization options for jsforce (see http://jsforce.github.io/jsforce/doc/Connection.html)
    
    SF_USER: process.env.SF_USER,
    SF_PASSWD_WITH_TOKEN: process.env.SF_PASSWD_WITH_TOKEN,
    
    REPORTPREFIX: "ReportOut_",
    // File name generated is REPORTPREFIX + reportid + startdate + enddate + execution timestamp
    
    OUTPUTTO: "file", 
    // This can be 'file' - to write results to a file; or 's3' - to write results to a S3 object.
    
    AWSCONFIG: {    
        accessKeyId: 'AKID', secretAccessKey: 'SECRET', region: 'us-west-2'
    }, 
    // This is required when you are using AWS S3 outside AWS Lambda and have not set the environment variables AWS_ACCESS_KEY and AWS_SECRET_KEY.  
    // See http://docs.aws.amazon.com/AWSJavaScriptSDK/latest/AWS/Config.html#constructor-property
        
    S3BUCKET: "", 
    // S3 bucket if OUTPUTTO is set to "s3".
    
    S3KEYPREFIX: "" 
    // S3 key prefix if OUTPUTTO is set to "s3". This is the path where you want to store the output file.
}

Using it in AWS Lambda

To run the downloader in AWS Lambda, you need to create a lambda zip package. If you have compiled node libraries, prepare this on a Linux machine.

• Prepare your environment.
  • Create an empty directory.
  • Install aws-sdk and jsforce_downloader.
• Create your lambda NodeJS script.
• Test your lambda function locally.
• Packaging the lambda function (creating a zip file for your AWS Lambda function).
• Create the Lambda function with the AWS CLI.
• Invoke the Lambda function with the AWS CLI.

Prepare your environment

mkdir myfunction
cd myfunction
npm install aws-sdk jsforce_downloader

Create your lambda NodeJS script.

Save this as index.js in the myfunction directory. Alternatively download index.js.

// AWS Lambda wrapper around JSFORCE_Downloader
'use strict';
var jsforce_downloader = require('jsforce_downloader');
/**
 * Event JSON should be
 * event.config = JSForce_Downloader config options that override the defaults.
 * event.options = JSForce_Downloader parameters to download the report.
 */
exports.handler = (event, context, callback) => {
        var config =  {
                WRITE_TEMP_FILES: false,
                REPORTPREFIX: "LambdaReportOut_",
                OUTPUTTO: "s3"
        };

        if (event.config) {
                for (var key in event.config)
                        config[key]=event.config[key];
        }
        var options = event.options; 
        if (options) {
                jsforce_downloader.initialize(config);
                jsforce_downloader.downloadreport(options.report, options.datefield,options.indexfield, options.startdate,options.enddate);
        }       
};

Source for your local test script.

• Save the following code as test.js in the myfunction directory. Alternatively download test.js.
• Edit the event to put in your Salesforce and AWS details. SF_USER, SF_PASSWD_WITH_TOKEN, S3BUCKET, S3KEYPREFIX.
• Edit the event to set the options.report, options.datefield, options.indexfield, options.startdate, options.enddate.

var index = require('./index.js');
// Create the event that will be passed to the handler.
var event =
{
        "config":{
                "MAX_CONCURRENT": 40,
                "WAIT_BETWEEN_REQUESTS":500,
                "REPORTSECTION": "T!T",
                "WRITE_TEMP_FILES": false,
                "SFOptions" : {
                        "loginUrl": "https://login.salesforce.com"
                },
                "AWSCONFIG": {    
                        "region": 'us-east-1'
                },
                "SF_USER" : "[e-mail]",
                "SF_PASSWD_WITH_TOKEN": "passwd_and_token",
                "REPORTPREFIX": "LambdaReportOut_",
                "OUTPUTTO": "s3",
                "S3BUCKET": "monima",
                "S3KEYPREFIX":"jsforce"

        },
        "options":{
                "report": "_salesforce_reportid_like_00OE0000002whwz",
                "datefield": "Case.CreatedDate",
                "indexfield": 0,
                "startdate":"2016-04-13",
                "enddate":"2016-04-15"
        }
};
index.handler(event);

Test your lambda function locally.

• Ensure the environment variables AWS_ACCESS_KEY and AWS_SECRET_KEY are set to your AWS credentials.
• Run node test.js to test the lambda function locally.
• Verify the file was successfully uploaded into S3.

$ node test.js
Starting here....
Report:00OE0000002whwz
Output to:LambdaReportOut_00OE0000002whwz_20160413-20160413_20160417220418.csv
Start:2016-04-13
End:2016-04-15
Logged into Salesforce
username: [e-mail](EI Heartbeat)
0:Start Range: (2016-04-13 to 2016-04-13)
1:Start Range: (2016-04-14 to 2016-04-14)
2:Start Range: (2016-04-15 to 2016-04-15)
1:Returned Range: (2016-04-14 to 2016-04-14) :Success:413 rows in section T!T
413 records
...
=============================
Report       :00OE0000002whwz
Date range   :2016-04-13 to 2016-04-15
Output to    :LambdaReportOut_00OE0000002whwz_20160413-20160415_20160417220418.csv
Done         :1232 records written.
Async reports:3 - (succeeded:3,failed:0).
Successfully uploaded data to monima/jsforce/LambdaReportOut_00OE0000002whwz_20160413-20160415_20160417220418.csv

Packaging the lambda function

• Setup your AWS CLI if you want to use the command line to create your lambda function. If you want to use the Web console, you don't need the AWS CLI. Instructions for CLI setup.

• Amazon has documented the process in their document "Creating a Deployment Package (Node.js)".

• From the myfunction directory, run the following on the command line. Alternatively download function.zip.

zip -rq function index.js node_modules README.md

• This will create a ZIP file for your lambda function and exclude the test.js file which has your credentials.

Create the Lambda function

• Note: You can use the Web console instead of the following steps.

• Get the ARN for the Lambda role "lambda_basic_execution" or "lambda_basic_execution_with_vpc". IAM Home. View the details of the role and copy down its ARN. It would look similar to arn:aws:iam::854421518417:role/lambda_basic_execution.

• Create a lambda function with 300 seconds timeout, upload the function.zip file. Amazon docs on creating a Lambda function. Remember to use the correct --profile user that you setup during the AWS CLI setup. If are using the default user, you can remove --profile adminuser from the command.

aws lambda create-function \
--region us-east-1 \
--function-name DownloadSFReport \
--zip-file fileb://./function.zip \
--role {role-arn} \
--handler index.handler \
--runtime nodejs4.3 \
--profile adminuser \
--timeout 300 \
--memory-size 1024

Invoke the Lambda function

• Copy the event JSON from your test.js and save it into event.json.
• Check that it is valid JSON (all keys and values are quoted). JSON Lint is a quick and easy way to check the validity.
• Invoke the function. Remember to use the correct function name if it is not DownloadSFReport. Use the correct --profile user that you setup during the AWS CLI setup. If are using the default user, you can remove --profile adminuser from the command.

aws lambda invoke \
--invocation-type RequestResponse \
--function-name DownloadSFReport \
--region us-east-1 \
--log-type Tail \
--payload file://./event.json \
--profile adminuser \
outputfile.txt

The preceding invoke command specifies RequestResponse as the invocation type, which returns a response immediately in response to the function execution. Alternatively, you can specify Event as the invocation type to invoke the function asynchronously. By specifying the --log-type parameter, the command also requests the tail end of the log produced by the function. The log data in the response is base64-encoded as shown in the following example response:

{
     "LogResult": "base64-encoded-log",
     "StatusCode": 200 
}

On Linux and Mac, you can use the base64 command to decode the log.

$ echo base64-encoded-log | base64 --decode

The following is a decoded version of an example log.

START RequestId: 231b8ce2-051c-11e6-84c3-af7b2d0cd02a Version: $LATEST
2016-04-18T04:15:10.683Z	231b8ce2-051c-11e6-84c3-af7b2d0cd02a	Starting here....
2016-04-18T04:15:10.683Z	231b8ce2-051c-11e6-84c3-af7b2d0cd02a	Report:00OE0000002whwz
2016-04-18T04:15:10.684Z	231b8ce2-051c-11e6-84c3-af7b2d0cd02a	Output to:LambdaReportOut_00OE0000002whwz_20160413-20160415_20160418040466.csv
...
2016-04-18T04:15:13.718Z	231b8ce2-051c-11e6-84c3-af7b2d0cd02a	Successfully uploaded data to s3://monima/jsforce/LambdaReportOut_00OE0000002whwz_20160413-20160415_20160418040466.csv
END RequestId: 231b8ce2-051c-11e6-84c3-af7b2d0cd02a
REPORT RequestId: 231b8ce2-051c-11e6-84c3-af7b2d0cd02a	Duration: 3218.25 ms	Billed Duration: 3300 ms 	Memory Size: 1024 MB	Max Memory Used: 79 MB

AWS errors and workarounds

• [PermanentRedirect: The bucket you are attempting to access must be addressed using the specified endpoint. You must specify a region to access your S3 bucket. Add this to your event.config or config.

    "AWSCONFIG": {
        "region": 'us-east-1'
    },

• Function was terminated or Function seems to be stuck. Lambda has a max timeout of 5 minutes and will terminate the function after that. Check the max memory used for your stuck function, and increase it if you are at the limit.

How it works

The library does the following

• Download the report metadata to setup the headers for the CSV file. I use the excellent node library csv-stringify to create CSV files.
• Add a date filter to the report metadata.
• Request execution of an Async report where the date filter is set to each day between the start and end dates. So if there are 365 days between the start and end date, it will generate 365 async reports.
• The Async reports are requested in sets of 30 each. This can be changed by setting MAX_CONCURRENT in the config.
• The program starts polling Salesforce to see if the Async reports are finished. The polling is done every 2000 ms. This can be changed by setting WAIT_BETWEEN_REQUESTS in the config. I don't recommend a number less than 500 ms.
• Download the results of the completed Async reports and store them in memory.
• When all async reports are completed, output to a file or to a S3 object.

Design choices

SF Async reports vs Sync reports: Async reports have a higher limit on the number of requests. This is important if you are downloading a lot of days. << TODO: What is the limit of ASync reports per hour per user? >>

Why this library?

I needed to automate the download of a large report to a CSV file. This task was done manually earlier and would take a long time to complete. So I looked into options using Node.

The excellent jsForce node module is a great wrapper around the Salesforce REST API. However it does not have a simple way to repeatedly call a report to get more than 2000 results. Unlike SOQL queries, there is no "queryMore" equivalent for reports. So I had to write a lot of non-trivial code to call the same report multiple times, switch to using asynchronous Salesforce reports, run multiple reports in parallel, etc.

After the first revision went out, I got requests to make this run in AWS Lambda and export the data to AWS S3 directly. This allows it to be a part of a AWS Data Pipeline to automate loading Salesforce report extracts into AWS RedShift.