dmreiland / sftp-fetch

Download new files from an SFTP server and publish them to a queue, using S3 to hold the data

sftp-fetch

Download new files from an SFTP server and publish them to a queue, using S3 to hold the data.

Configuration

First, create a properties file that looks like

sftp.hostname=<the.sftp.hostname>
sftp.username=<the.sftp.username>
sftp.password=<the.sftp.password>
sftp.folder=<the.sftp.folder.to.download.from>

rabbit.hostname=<the.rabbit.hostname>
rabbit.port=5672
rabbit.username=<the.rabbit.username>
rabbit.password=<the.rabbit.password>
rabbit.vhost=<the.rabbit.vhost>
rabbit.exchange=<the.rabbit.exchange>

s3.bucket=<s3.bucket>
fetch.days=<number.of.days.to.fetch>

Usage

The Basics

Once you have created the appropriate config files you can invoke

java -jar sftp-fetch.jar -c </path/to/properties/file> -n

This will show you what files it would process without actually doing it. Run without -n to actually process. Files that have already been processed will be skipped unless you run with --overwrite

Files that match a pattern

Optionally you can restrict operations only to files that match a certain pattern using -p

java -jar sftp-fetch.jar -c </path/to/properties/file> -r <routingkey> -p '.*SomeFileName.*'

Specify the routing key

You can specify a routing key using --routing-key at the command line or rabbit.routingkey in the property file.

GPG decryption

If you are fetching GPG encrypted files from SFTP you can optionally have sft-fetch decrypt them by supplying the path to the GPG private key in the property file using

decryption.key.path=</path/to/pgp/private/key>

Currently encrypted private keys are not supported.