Feed FAQs

Where is my uploaded feed file?

Once a file is uploaded, Personalization Platform begins processing it immediately, so expect it to disappear from the FTP directory. This ensures that your catalog feed is processed as soon as possible.

Will temporary file names cause problems?

Uploading a file with a temporary name and then renaming it will prevent the feed from processing.

It is advised to use client software that does not rename files. Upload the file using the final file name. You may have to change your FTP client or change your process to not use a temporary file name.

What compression types are supported?

  • For feeds containing multiple files: zip and tar.gz

  • For single files: plain text or .gz

Please zip or tar.gz feeds that contain multiple files.

Will special characters cause any issues?

We support UTF-8 encoding, which can theoretically encode any character. Typically, character issues are the result of using the incorrect FTP transfer method when a customer uses Windows systems (given that Personalization Platform uses Linux). Binary files (including (g)zip files and tarballs) should be transferred in binary mode, and text files (such as straight .xml or .txt files) should be transferred in ASCII mode.

If you are still using the deprecated XML feed: It is possible to specify special characters for attribute names in the XML feed, but they must conform to a specific format. That format is &#xxxx; where xxxx represents the 4-digit hexadecimal Unicode code point of the special character that is being referenced. For example, if you would like to deliver an attribute named François, the name in the XML feed would look like this: François

Are there forbidden characters in the catalog data feed?

If it is an XML format, reserved XML characters still apply. Whatever character is being used as the column delimiter in the flat feed format cannot be escaped. In all cases, only UTF-8 encoding is supported.

Why is a 0-byte file showing in my FTP client?

Though the file has a name, it does not contain content. This may be because your firewall didn’t support passive mode FTP upload. In this case, force your FTP client to use passive mode.

Can I use the archive directory?

No. The archive directory is for the Algonomy Operations team use only. Files should not be uploaded here. If a feed is compressed, our process may use this directory to decompress the files. Therefore, a copy of this file may be found in this archive directory during or after processing. The archive directory may contain a backup of the decompressed file.

What about permissions when uploading the file?

The file should be uploaded with its permissions set at 664.

Is SFTP supported?

Yes, SFTP is supported for multiple data ingestion workflows in the Personalization Platform, including catalog feed uploads and offline batch ingestion.

For catalog feeds, the compressed feed file (such as .zip or .gz) is uploaded to the Algonomy SFTP server. This is the standard and recommended method for securely transferring catalog data on a scheduled basis.

In addition to catalog feeds, SFTP is also supported for offline ingestion workflows such as UPSBATCH, UPSLINK, and HIVEUPLOAD, where files are uploaded along with metadata and processed asynchronously.

We also support FTPS (FTP over SSL) for secure FTP-based transfers. Most FTP clients support FTPS out of the box, and the default port for implicit FTPS is 990.

For delivery of feeds to an ESP (email service provider), both FTP and SFTP are supported.