What process fills the disk every 2 hours?

Support
1

I have a site with a 100 GB database on a 160GB SSD. Images on S3.

Every two hours postgres does something that fills the disk. It’s not a backup.

I guess I should just up the disk to 200GB?

4 Likes
2

Now the disk has 320GB, so there is no way that the disk is getting full, but still every two hours I get this:

Job exception: PG::DiskFull: ERROR: could not resize shared memory segment "/PostgreSQL.2431023752" to 134217728 bytes: No space left on device CONTEXT: parallel worker
4:31 am
17
Job exception: Failed to open TCP connection to 84.247.3.95:80 (Connection refused - connect(2) for "84.247.3.95" port 80)
5:56 am
Job exception: PG::DiskFull: ERROR: could not resize shared memory segment "/PostgreSQL.4192722046" to 134217728 bytes: No space left on device CONTEXT: parallel worker
6:26 am
Job exception: PG::DiskFull: ERROR: could not resize shared memory segment "/PostgreSQL.175048658" to 134217728 bytes: No space left on device
8:32 am
3
MD Composer Extras theme/component is throwing errors: TypeError: Cannot read properties of null (reading 'es')
8:36 am
86
Unicorn worker received USR2 signal indicating it is about to timeout, dumping backtrace for main thread config/unicorn.conf.rb:203:in `backtrace' config/unicorn.conf.rb:203:in `block (2 levels) in re
8:50 am
Job exception: PG::DiskFull: ERROR: could not resize shared memory segment "/PostgreSQL.2657529202" to 134217728 bytes: No space left on device CONTEXT: parallel worker

image
image1217×207 44 KB

3

stackoverflow suggests

This is because docker by-default restrict size of shared memory to 64MB.

You can override this default value by using --shm-size option in docker run.

2 Likes