How to clean an 'updated' nginx proxy manager LXC?

[samwathegreat]

Hey guys,

I have an nginx proxy manager LXC that I've updated a few times. It always seems to work great, but each update increases the disk usage about a gigabyte or so. I've had to resize the disk a couple times now and it's approaching capacity again.

Does anyone know why it keeps increasing in space and never re-claiming any space? It's like it keeps adding and not replacing during each update... Is there a way to 'clean' it? 'apt autoremove' does nothing...

I really hate to start 'fresh' with a new lxc because there's no backup/restore functionality (why in the world there isn't is beyond me) and I have quite a few hosts that I'd prefer not to reconfigure.

Thanks in advance!

[MickLesk]

That are Node Caches (f.e. yarn, pnpm, npm) you can run f.e. yarn cache clean with yarn tools to remove the build Cache. We have removed this cleaning in our core, because some apps constantly broke after cleaning. You can try it by your own, but i recommend to do an snapshot or backup before.

[drVLecter]

Hi @samwathegreat

Ill paste below what I usually do and Ill try to make it a bit lxc agnostic so you can use it for any lxc, not just NPM:

Housekeeping:

1. Backup! Make a full lxc backup because you will be deleting stuff and if the service fails you can just restore from backup.

Checkup:

2. Run these commands:

df -h
du -h --max-depth=1 / 2>/dev/null | sort -hr | head -20

du -sh /var/cache/apt/
du -sh /var/log/

First two will show space used and what folders are the highest space users. Next two shows you how big are apt caches and logs. There are more commands that you can run but these will start giving you a sense how do go about this process.

Cleanup:

3. Run these commands:

apt clean && apt autoclean
apt autoremove --purge
journalctl --vacuum-time=7d
pnpm store prune
npm cache clean --force
yarn cache clean

First two will cleanup previous updates leftovers; third one will delete all logs older than 7 days - Im assuming if the service is running well you dont need old logs; next 3 will remove old node caches whether it relies on pnpm, npm or yarn - if it doesnt it will just say - not found.

That is the most basic setup - you can investigate more folders using the Checkup section by editing those commands. Just never delete node modules or service main folder and always backup before making cuts.

[samwathegreat]

Thanks, Gentlemen

These two commands reclaimed 75% of my storage for this particular lxc!

npm cache clean --force
yarn cache clean

(the pnpm command didn't work in this lxc - 'command not found').

It doesn't seem to have broken anything on the NPM lxc, but I have a snapshot & backup to fall back to just in case there is an issue at the next update.

Appreciate the help.

[drVLecter]

that's good mate.

I made a list of commands that should* work on any lxc - not just NPM.
And depending on the backend at least one or two of those last commands will just respond with - not found.

Just remember to always backup before and test after to make sure it works as expected.