[LU-1375] Are LUNs over 24T safe in Lustre 1.8.7?

• Status: Resolved
• Project: Lustre
• Type: Bug
• Priority: Minor
• Reporter: Roger Spellman (Inactive)
• Assignee: Peter Jones
• Resolution: Fixed

Environment

I am using the 1.8.7 kernel with RHEL 5.4. The RPMs I have are:

e2fsprogs-1.41.90.wc3-0redhat
e2fsprogs-devel-1.41.90.wc3-0redhat
lustre-1.8.7-wc1_2.6.18_274.3.1.el5_lustre.g9500ebf

• Severity: 3

Description

I am using the 1.8.7 kernel with RHEL 5.4. The RPMs I have are:

e2fsprogs-1.41.90.wc3-0redhat
e2fsprogs-devel-1.41.90.wc3-0redhat
lustre-1.8.7-wc1_2.6.18_274.3.1.el5_lustre.g9500ebf

When I try to build a OST greater than 24TB, it fails, with the following message in dmesg:

LDISKFS-fs does not support filesystems greater than 24TB and can cause data corruption.Use "force_over_24tb" mount option to override.

That sounds pretty bad.

How risky is it to use a LUN greater than 24TB?

Are there some known issues with it? If so, what are they?

In any case, where does this parameter go in the following command line:

mkfs.lustre --reformat --ost --mkfsoptions="-m 0 " --fsname big --mgsnode=${MGSNODE} /dev/mapper/map00

?

I tried using --force_over_24tb as an option to mkfs.lustre. This gave me:

mkfs.lustre: unrecognized option `--force_over_24tb'

I tried -O force_over_24tb added to mkfsoptions. This gave me:

Warning: the fs_type huge is not defined in mke2fs.conf
Invalid filesystem option set: force_over_24tb,extents,uninit_bg,dir_nlink,huge_file,64bit,flex_bg

Thanks.

Comments

[Peter Jones]

Roger

We have only tested up to 24TB LUNs for 1.8.x releases. If you need larger LUN sizes it would be recommended to use a 2.1.x release

Regards

Peter