Support

EXTREMELY IMPORTANT: Please attach a ZIP file containing your Akeeba Backup log file in order for us to help you with any backup or restoration issue. If the file is over 2Mb, please upload it on your server and post a link to it.

Description of my issue:

I've been trying out uploading my backups to Amazon S3 and that works fine. But I then download the backup from S3 to my computer so I can use Akeeba Extract wizard to unpack it. When I do that I get this error:

A decompression error occured while extracting the archive.

I did some searching and that error is apparently caused by a part missing in a split archive backup. I have looked through and made sure all 223 parts are there so nothing is missing. When I try to do the exact same but with a .jpa backup it works fine. It's just with .jps file I can't unpack my backup.

Could it be caused by a part of the backup being corrupt or not complete?

A missing or corrupt part of the archive would cause this error. All of the parts of the archive should be the same size except the last (jps) part.

We recommend that you download (and upload) your archive via FTP in binary mode. Binary mode is important because ASCII mode will corrupt your archive and Auto mode makes your FTP client guess which mode it should use.

Yes all the parts are exactly the same size.

I'm using Cloudberry to download the backup directly from S3 to my computer. So I'm not using FTP. And it did work with a .jpa backup so that would indicate that Cloudberry downloads it the right way, is that a correct assumption?

I now tried to instead saving the jps archive on my own server without uploading it to S3 and then with FTP downloading it to my computer. Yes I downloaded it in binary mode.

And I still get the same decompression error when I try to extract it. But I don't get that when I use a jpa archive.

Any idea why that is?

Is the problem with a single encrypted archive or several - is the problem consistent?

Are you sure you have the right password? I know I'm a decent typist except when I'm typing a password, then I have eleven thumbs. :-)

There is no way to check the password that you entered in the Configuration page, it is encrypted when it is stored. You can re-enter what you wanted the password to be just to be double sure you don't have a typo in that field.

Is your password limited to a-z, A-Z and 0-9 characters? In theory other characters should work but they are a gray area.

Yes I put it in again to be sure and I copy pasted it.

I now also tried a one part archive, so not a split on. And I downloaded it with FTP to my computer but I still get the same decompression error at I think the same place. It always seems to be extracting a image folder when that happens.

Is there anything else I could try to find out what the problem is?

I now tried to remove the image it got stuck on and then the decompression got a bit further but I still ended up with the same error messages. This time it stopped on a flv file. Strange thing it works fine with .jpa archives but not .jps.

It can't be the data because if it was, jpa wouldn't work either. It just doesn't make any sense.

I'm going to ask Nicholas to look at this.

That was the problem, Kickstart worked fine!

Another question, is it possible to make Akeeba store backups in folders on Amazon S3? So that it creats one folder for each backup. Would make it a lot easier to browse through manually.

You can set the folder name using one of the naming variables listed here: https://www.akeebabackup.com/documentation/akeeba-backup-documentation/configuration.html#basic-configuration. The [DATE] and [TIME] variables work well for this. Note that in some circumstances either of those variables can put your backup into multiple folders if the date or time variable changes during your backup.

That works fine when I save backups on my server but when I upload them to S3 the directory isn't created. Is it possible to put the backup in a directory on S3 also?

We're talking about the directory field in the S3 setup, my setting is:

/mysite.com/[DATE]-[TIME]

And S3 creates the folder just fine. Make sure the user credentials you are using have permission to create the folder.