Skip to main content
FlashGrid.io

Understanding low or negative Usable_file_MB values

  • Updated
  • Products: FlashGrid Cluster on Azure
  • Products: FlashGrid Cluster on GCP
  • Products: FlashGrid Storage Fabric
  • Products: FlashGrid Cluster on AWS

Note: This article was revised in August 2023 and provides updated guidance from FlashGrid.

Negative Usable_file_MB

FlashGrid strongly recommends adding more disks to the disk group when Usable_file_MB approaches zero or immediately after it becomes negative.

Usable_file_MB value for an ASM disk group indicates the total size of files that can be safely added to the disk group.

Usable_file_MB can become negative in asmcmd lsdg output and in V$ASM_DISKGROUP and V$ASM_DISKGROUP_STAT views

Example:

$ asmcmd lsdg data
State   Type  Rebal  Sector  Logical_Sector  Block       AU  Total_MB  Free_MB  Req_mir_free_MB  Usable_file_MB  Offline_disks  Voting_files  Name
MOUNTED HIGH  N         512             512   4096  4194304 122880000 10088316         15360000        -1757228              0  N             DATA/

The disk group is likely to continue working and it may be possible to create new files on the disk group even after Usable_file_MB becomes negative. However, the following problems may start happening:

  • ASM may have insufficient free space to reliably perform its internal operations, such as rebalance.
  • Excessive imbalance of free space between disks may start appearing.
  • Creating new files or extending existing files may result in errors even when Free_MB is large enough to accommodate those files.

Best practices for managing disk group space

To avoid excessive amount of space held as reserved, FlashGrid recommends configuring either 1 disk per node for small capacity disk groups, or 5+ disks per node for large capacity disk groups.

A FlashGrid Cluster typically has one of the following two configurations:

  • Two database nodes: Normal Redundancy with two regular failgroups (one per db node) plus one quorum failgroup
  • Three database nodes: High Redundancy with three regular failgroups (one per db node) plus two quorum failgroups

In these configurations capacity of a single disk per failgroup is kept in “reserve”, except where there is only one disk per failgroup. The Usable_file_MB value is reduced by the amount of the “reserved” space.

For disk groups with more than one disk per failgroup, it is important to pay attention to the percentage of the total raw capacity that will be reserved. Too few disks (e.g, 2, 3, or 4) will result in significant percentage of the raw space allocated for the reserve, which is undesirable for cost reasons.

Amount of the “reserved” space as percentage of the total raw space:

  • 1 disk: 0% (no space is reserved with one disk per failgroup)
  • 2 disks: 50%
  • 3 disks: 33%
  • 4 disks: 25%
  • 5 disks: 20%
  • 10 disks: 10%
  • 20 disks: 5%

Oracle Enterprise Manager usage notes

The following Oracle Enterprise Manager (OEM) metrics use Usable_file_MB and can trigger alerts:

  • Disk Group Usable (Free MB)
  • Disk Group Usable (MB)
  • Disk Group Used % of Safely Usable

The metrics are in the Disk Group Usage metric category and apply to Cluster ASM targets. FlashGrid recommends using these alerts and expanding disk group capacity as needed to avoid negative Usable_file_MB.