Bug #56450
Updated by Xiubo Li almost 2 years ago
Before running the Filesystem benchmark test, from OS we can see that the *_/_* directory had *_81GB_*: <pre> [root@lxbceph1 kcephfs]# df -h Filesystem Size Used Avail Use% Mounted on devtmpfs 14G 4.0K 14G 1% /dev tmpfs 14G 84K 14G 1% /dev/shm tmpfs 14G 50M 14G 1% /run tmpfs 14G 0 14G 0% /sys/fs/cgroup /dev/mapper/rhel-root 235G 155G 81G 66% / /dev/sda1 15G 1.4G 14G 9% /boot /dev/loop1 39M 39M 0 100% /var/lib/snapd/snap/gh/545 /dev/loop0 39M 39M 0 100% /var/lib/snapd/snap/gh/544 /dev/loop4 56M 56M 0 100% /var/lib/snapd/snap/core18/2409 /dev/loop2 45M 45M 0 100% /var/lib/snapd/snap/snapd/15904 /dev/loop3 56M 56M 0 100% /var/lib/snapd/snap/core18/1944 tmpfs 2.8G 16K 2.8G 1% /run/user/42 tmpfs 2.8G 0 2.8G 0% /run/user/0 10.72.47.117:40545,10.72.47.117:40547,10.72.47.117:40549:/ 99G 0 99G 0% /mnt/kcephfs </pre> And after the test finished, it had has *_26GB_*: <pre> [root@lxbceph1 usr]# df -h Filesystem Size Used Avail Use% Mounted on devtmpfs 14G 4.0K 14G 1% /dev tmpfs 14G 84K 14G 1% /dev/shm tmpfs 14G 50M 14G 1% /run tmpfs 14G 0 14G 0% /sys/fs/cgroup /dev/mapper/rhel-root 235G 210G 26G 90% / /dev/sda1 15G 1.4G 14G 9% /boot /dev/loop1 39M 39M 0 100% /var/lib/snapd/snap/gh/545 /dev/loop0 39M 39M 0 100% /var/lib/snapd/snap/gh/544 /dev/loop4 56M 56M 0 100% /var/lib/snapd/snap/core18/2409 /dev/loop2 45M 45M 0 100% /var/lib/snapd/snap/snapd/15904 /dev/loop3 56M 56M 0 100% /var/lib/snapd/snap/core18/1944 tmpfs 2.8G 16K 2.8G 1% /run/user/42 tmpfs 2.8G 0 2.8G 0% /run/user/0 10.72.47.117:40545,10.72.47.117:40547,10.72.47.117:40549:/ 99G 0 99G 0% /mnt/kcephfs </pre> And from *_ceph df_* we can see that the spaces had been released: <pre> [root@lxbceph1 build]# ./bin/ceph df --- RAW STORAGE --- CLASS SIZE AVAIL USED RAW USED %RAW USED hdd 303 GiB 300 GiB 3.5 GiB 3.5 GiB 1.14 TOTAL 303 GiB 300 GiB 3.5 GiB 3.5 GiB 1.14 --- POOLS --- POOL ID PGS STORED OBJECTS USED %USED MAX AVAIL .mgr 1 1 577 KiB 2 1.7 MiB 0 99 GiB cephfs.a.meta 2 16 156 MiB 61 468 MiB 0.15 99 GiB cephfs.a.data 3 32 0 B 0 0 B 0 99 GiB </pre> If I run the same test again it will fail with no disk spaces and from the *_df -h_* we can see that the disk spaces from OS has been used up: <pre> [root@lxbceph1 usr]# df -h Filesystem Size Used Avail Use% Mounted on devtmpfs 14G 4.0K 14G 1% /dev tmpfs 14G 84K 14G 1% /dev/shm tmpfs 14G 50M 14G 1% /run tmpfs 14G 0 14G 0% /sys/fs/cgroup /dev/mapper/rhel-root 235G 235G 103M 100% / /dev/sda1 15G 1.4G 14G 9% /boot /dev/loop1 39M 39M 0 100% /var/lib/snapd/snap/gh/545 /dev/loop0 39M 39M 0 100% /var/lib/snapd/snap/gh/544 /dev/loop4 56M 56M 0 100% /var/lib/snapd/snap/core18/2409 /dev/loop2 45M 45M 0 100% /var/lib/snapd/snap/snapd/15904 /dev/loop3 56M 56M 0 100% /var/lib/snapd/snap/core18/1944 tmpfs 2.8G 16K 2.8G 1% /run/user/42 tmpfs 2.8G 0 2.8G 0% /run/user/0 10.72.47.117:40545,10.72.47.117:40547,10.72.47.117:40549:/ 99G 8.5G 91G 9% /mnt/kcephfs </pre> Is this a bug ? If not how to fix this ?