jira directory structure

[arno]

Hi All

We recently found no storage server, is considering expanding or clean up the file processing, we found that under the log path of the data folder (var/atlassian/application data/jira/data) take up larger, want to know which store what is the main data? Is there any relevant documentation for reference? I haven't found it in the official documentation yet

my jira version is 8.13.3

[Nic Brough -Adaptavist-]

Jira keeps its data in two places, but there is a third to worry a bit about.

• Jira installation is where the software is installed and runs from.  You may want to look at the logs in the installation, but generally, this one won't grow, and a default install cleans out the logs regularly anyway.
• The core Jira data (issues, config, etc) lives in a database.  Most databases do some compression and encoding, and you'll find their disk space grows quite slowly (I saw a Jira with over a million issues in it, and the database server hadn't got close to the 100Gb disk it was on)
• "Jira home" is the phrase most commonly used to describe Jira's working space - it's used for all sorts of data storage - temporary files, plugin data, logs etc.  But this can, and often does, grow with usage.  The bits to look at in there are:
  • Logs again (but again, the default install usually puts in some log rotation)
  • Automatic XML backups in the export directory - the installer does not try to control these; they just grow constantly.  The default is to backup twice a day, but these backups are usually only useful to small systems; most of us turn them off, and backup the database directly instead, as well as the attachments
  • Attachments are the usual culprit.  Every time someone attaches a file to an issue, the file gets added to the disk, uncompressed, and they're only deleted when you delete (not just archive) the issue to which they are attached.

In short, look at the server logging, backups and attachments!

[arno]

Hi @Nic Brough -Adaptavist- 

Thank you for your reply

As you said, I have checked the files in the data directory and found that the attachments occupy a large space, some of which are historical data. We hope to reduce the occupation of historical data (such as migrating to a new machine), but we do not want to affect the use of users. Besides expansion, is there any better way to help us achieve optimization

[Nic Brough -Adaptavist-]

The only way to recover disk space from attachments is to delete attachments, best done by deleting issues. (You could delete them from the disk, but then you just get an unclickable little "broken file" icon in your Jira issues)

If you need to keep them, then your only option is to increase disk space.