zos icon indicating copy to clipboard operation
zos copied to clipboard

Mounted Volume in a Container shows 100 % usage

Open abdulgig opened this issue 5 years ago • 0 comments

Overview

Link to reservation : https://explorer.testnet.grid.tf/explorer/reservations/460837

We have a 2000 GB volume attached to the container. 11TB is the HDD on the node. The volume is writable, we can download files, no issues there ! See below,

:~# df -h
Filesystem      Size  Used Avail Use% Mounted on
overlay         448G   76M  448G   1% /
tmpfs            64M     0   64M   0% /dev
shm              64M     0   64M   0% /dev/shm
tmpfs           1.5G  971M  566M  64% /etc/resolv.conf
/dev/sdb         11T  1.7G     0 100% /dgb-data
tmpfs            63G     0   63G   0% /proc/acpi
tmpfs            63G     0   63G   0% /sys/firmware
tmpfs            63G     0   63G   0% /proc/scsi

Problem

The problem is that the client binary/software sees it as 100 % used and therefore aborts with the error that Disk space is low. Here is it,

~# ./digibyte/bin/digibyted --datadir=/dgb-data
2020-05-22T13:50:30Z Checking all blk files are present...
2020-05-22T13:50:30Z Initializing databases...
2020-05-22T13:50:30Z *** Disk space is low!
2020-05-22T13:50:30Z Error: Error: Disk space is low!
Error: Error: Disk space is low!
2020-05-22T13:50:30Z ERROR: out of disk space

We already know that btrfs subvolumes won't be able to show the correct used/free values inside the container, which is fine ! We need to understand why does it show 100 % and why not 0.1 % ?

/dev/sdb 11T 1.7G 0 100% /dgb-data

Access to Container

ZOS NodeID : 3C2d8LaXAbEwEpdnDTx7HhhtMZhTmYJzSw4EdRMuMFfC
ZOS node IPv6 address : 2600:70ff:d0d3:0:225:90ff:fe4f:629e

How to test ?

Log in with the credentials above,

tmux a -t space
./digibyte/bin/digibyted --datadir=/dgb-data

abdulgig avatar May 22 '20 14:05 abdulgig