No space left on device error (even though there's plenty)

First, thanks for mup, it’s an awesome tool!

I’m deploying on EC2 and I’ve been getting “no disk space” errors recently. But when I SSH into the box and run df -h, I see:

Filesystem      Size  Used Avail Use% Mounted on
udev            484M   12K  484M   1% /dev
tmpfs           100M  572K   99M   1% /run
/dev/xvda1       24G   13G  9.7G  57% /
none            4.0K     0  4.0K   0% /sys/fs/cgroup
none            5.0M     0  5.0M   0% /run/lock
none            497M  476K  496M   1% /run/shm
none            100M     0  100M   0% /run/user

So there should be enough space (I’d think 9.7 GB is plenty).

This is the error I’m getting:

Error processing tar file(exit status 1): open /programs/server/npm/node_modules/react-router/node_modules/create-react-class/node_modules/fbjs/lib/someSet.js.flow: no space left on device
	-----------------------------------STDOUT-----------------------------------
	base: Pulling from abernix/meteord
	Digest: sha256:680d55194f4cce47dac3373ad54d7c3ca53e28e7c8f40d29ef43f82fa460da4b
	Status: Image is up to date for abernix/meteord:base
	sharewaste
	Sending build context to Docker daemon  256.6MB


	Step 1/15 : FROM abernix/meteord:base
	# Executing 1 build trigger...
	Step 1/1 : ENV NODE_VERSION ${NODE_VERSION:-4.8.3}
	 ---> Using cache
	 ---> 95e2bed2dd09
	Step 2/15 : RUN mkdir /built_app
	 ---> Using cache
	 ---> 15751d26d2dd
	Step 3/15 : COPY ./ /built_app
	sharewaste

Any ideas please?

[realyze]

Sorry guys, I lost track of this. My inode count was pretty low so I guess that was it. I have updated docker to 17.11 and ran docker system prune -a and that helped massively.

[nzomedia]

If there’s enough space on disk you could check to see if there’s still free inodes with df -ih source: https://unix.stackexchange.com/questions/40218/tar-no-space-left-on-device