Jira is reporting that we have a free space problem. Low disk space on device containing temporary

[kedar]

Jira is reporting that we have a free space problem.  Low disk space on device containing temporary directory (/opt/atlassian/jira/temp) - 6 GB of free space remaining. Index backup and recovery may fail

[Matt Doar]

And, is it so?

[kedar]

how i can reduce the space in (/opt/atlassian/jira/temp) 

[Matt Doar]

Stop Jira, remove files in that directory. But it's not common to have problems there. Make sure you have enough space in /opt too. 

[Sagar Mahajan]

This issue is also occur when the health check is not completed within the default time which is 2000 ms and this can be something related to the disk I/O performance as well.

You can increase the health check time out as mentioned in the below document, if the issue still persists then check the disk I/O and used the SSD which has higher throughput and disk I/O

JIRA Health Check Timeout 

[Nic Brough -Adaptavist-]

You are going to need to do some housekeeping if you can't extend the disk space.

Start with working out what partition the directory is in.  /opt is usually either on the root drive or with its own partition, but there's no way we can know if anyone has done clever things with your file system (such as mount things lower in the tree.  I often mount attachment storage within the application home directory for example)

Once you know what partition it is one, you need to work out where the space has gone.  Lets say /opt is in it's own partition, then you can, as root, run

cd /opt

du -kh -d1

This will list the size of each directory in there, so you can see what is chewing up all the space.  You can descend into each large one and repeat the du command for more detail.

[Mike Wise]

I am seeing this error too but only have used 2% of the file system where the temp directory is and only contains 12K of files. 

Something here is reporting falsely. 

[Matt Doar]

https://jira.atlassian.com/browse/JRASERVER-65802 now tracked in https://ecosystem.atlassian.net/browse/ATST-812

Yes, there is a bug in the library used by Jira to calculate disk space. If it cannot find a Lucene index file due to merging then it reports it as an error. Which is treated as running out of disk space, no matter how much space you have.

We have disabled 

Atlassian JIRA - Plugins - Instance Health Plugin 2.1.21.01

jira.healthcheck.storage.index.name

 (storageIndexSnapshotHealthCheck)

in Jira 7.2.13