Support

Description of my issue:

When I back up using Akeeba I have no indication that anything is wrong, it looks like a successful backup.

I carefully transfer my .jpa file in BINARY mode and check the file sizes on the server and file destination, they match.

When I attempt to restore using Kickstart, I get invalid header error.

When I attempt to use Akeeba eXtractor to extract the files locally, I get the archive format error midway through the extract. I can see some files/folders were extracted, not others.

It seems that Akeeba is not correctly archiving to .jpa, and/or eXtractor is not properly extracting. Why?

I've been using Akeeba to both backup and restore this site for a while now. Restores used to work fine. Then Akeeba started failing with the archive format error and I had to resort to Akeeba eXtractor to restore. Now even that isn't working. Additionally, old .jpa archives that I've actually successfully restored to another site now no longer work, .jpa's going all the way back to the original site give the same error - even though I have at least one other site on a different server which was restored normally with Akeeba and non-normally with eXtractor. No, there haven't been any major changes to the site. Has there been a change to Backup/eXtractor?

One thing that will cause the Invalid Header error is a file that changes size while it is being archived. When this happens, the header has one file size while the actual file size is different. It can happen with a log file, for example.

Run the ALICE log analyzer on your backup log. One of the things that ALICE will look for is common log file names that we know can cause problems. If ALICE finds a problem file, we can probably exclude it with a regular expression.