Pool have no space and some osd over 95%

rahulraj
19 Posts
October 20, 2025, 5:16 amQuote from rahulraj on October 20, 2025, 5:16 amwe have total 263TB storage space in petasan cluster. we have only nfs and iscsi shares. nfs has 70TB and iscsi has 50TB ,total 120TB used. but in ui dashboard shows 215/263 TB used. earlier we have some iscsi and nfs shares , but i deleted that shares but no changes happened in used space. some osds show 94%,95%,93%. i have doubted any deleted shares running in background or exist.
we have total 263TB storage space in petasan cluster. we have only nfs and iscsi shares. nfs has 70TB and iscsi has 50TB ,total 120TB used. but in ui dashboard shows 215/263 TB used. earlier we have some iscsi and nfs shares , but i deleted that shares but no changes happened in used space. some osds show 94%,95%,93%. i have doubted any deleted shares running in background or exist.

admin
3,054 Posts
October 20, 2025, 6:29 pmQuote from admin on October 20, 2025, 6:29 pmThe quickest way is to delete any un-needed data, or add OSDs.
The 120 TB used, is this before replication, if so then the used storage should be x3 this
Deleting from NFS should reflect space freed, from iSCSI deleting files will not reclaim space unless you enable trim/discard on the iSCSI drives and issue trim request from the guest OS.
You should also have the balancer enabled from UI to make sure the OSD usage is balanced.
The quickest way is to delete any un-needed data, or add OSDs.
The 120 TB used, is this before replication, if so then the used storage should be x3 this
Deleting from NFS should reflect space freed, from iSCSI deleting files will not reclaim space unless you enable trim/discard on the iSCSI drives and issue trim request from the guest OS.
You should also have the balancer enabled from UI to make sure the OSD usage is balanced.
Pool have no space and some osd over 95%
rahulraj
19 Posts
Quote from rahulraj on October 20, 2025, 5:16 amwe have total 263TB storage space in petasan cluster. we have only nfs and iscsi shares. nfs has 70TB and iscsi has 50TB ,total 120TB used. but in ui dashboard shows 215/263 TB used. earlier we have some iscsi and nfs shares , but i deleted that shares but no changes happened in used space. some osds show 94%,95%,93%. i have doubted any deleted shares running in background or exist.
we have total 263TB storage space in petasan cluster. we have only nfs and iscsi shares. nfs has 70TB and iscsi has 50TB ,total 120TB used. but in ui dashboard shows 215/263 TB used. earlier we have some iscsi and nfs shares , but i deleted that shares but no changes happened in used space. some osds show 94%,95%,93%. i have doubted any deleted shares running in background or exist.
admin
3,054 Posts
Quote from admin on October 20, 2025, 6:29 pmThe quickest way is to delete any un-needed data, or add OSDs.
The 120 TB used, is this before replication, if so then the used storage should be x3 this
Deleting from NFS should reflect space freed, from iSCSI deleting files will not reclaim space unless you enable trim/discard on the iSCSI drives and issue trim request from the guest OS.
You should also have the balancer enabled from UI to make sure the OSD usage is balanced.
The quickest way is to delete any un-needed data, or add OSDs.
The 120 TB used, is this before replication, if so then the used storage should be x3 this
Deleting from NFS should reflect space freed, from iSCSI deleting files will not reclaim space unless you enable trim/discard on the iSCSI drives and issue trim request from the guest OS.
You should also have the balancer enabled from UI to make sure the OSD usage is balanced.