[LU-8974] Сhange force_over_256tb lustre mount option to force_over_512tb for ldiskfs

• Status: Resolved
• Project: Lustre
• Fix Version/s: Lustre 2.10.0
• Type: Improvement
• Priority: Minor
• Reporter: Artem Blagodarenko (Inactive)
• Assignee: Andreas Dilger
• Resolution: Fixed

Description

Currently attempts of creating ldisk file system with size >256TB finished with message.
"LDISKFS-fs does not support file systems greater than 256TB and can cause data corruption.Use "force_over_256tb" mount option to override."
Before using "force_over_256tb" parameter in production systems lustre file system software should be analyzed to point possible large disks support issues. This issue is about research of some aspects of Lustre software. Finally patch that change "force_over_256tb" to "force_over_512tb" should be landed. This gives ability use ldiskfs partitions <512tb without options.

Issue Links

• is related to: LU-8443 Can't create Lustre FS with enabled m...
• is related to: LU-1365 Implement ldiskfs LARGEDIR support fo...
• is related to: LU-12950 Сhange force_over_512tb lustre mount ...
• is related to: LU-8444 Add test for LU-7325 patch
• is related to: LU-8976 Apply patch "libext2fs: fix maximum b...
• is related to: LU-7592 Сhange force_over_128tb lustre mount ...
• is related to: LU-8465 parallel e2fsck performance at scale

Comments

[Gerrit Updater]

Artem Blagodarenko ([e-mail]) uploaded a new patch: https://review.whamcloud.com/24524
Subject: LU-8974 osd-ldiskfs: increase supported ldiskfs fs size
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: b203e05a88d6c02b7f655598aacabfeb946aed66

[Artem Blagodarenko (Inactive)]

Some researched aspects

META_BG
(required)
Without the option META_BG, for safety concerns, all block group descriptors copies are kept in the first block group. Given the default 128MiB(2^27 bytes) block group size and 64-byte group descriptors, ext4 can have at most 2^27/64 = 2^21 block groups. This limits the entire filesystem size to 2^21 ∗ 2^27 = 2^48 bytes or 256TiB.
The solution to this problem is to use the metablock group feature (META_BG), which is already in ext3 for all 2.6 releases. Some issues found during testing
LU-8443 "Can't create Lustre FS with enabled meta_bg option". Patch is landed.
LU-8976 "Cannot create filesystem with requested number of inodes while setting up superblock" Patch from upstream must be cherrypicked to solve the issue.

large dir
(useful in case of many files in the directory)
LU-1365 "Implement ldiskfs LARGEDIR support for e2fsprogs" Sending upstream.

xattr external inode referance
(tested, landing for regression testing)
LU-8444 the patch from LU-7325 fixes this issue: "Hitting "LDISKFS-fs error (device md66): ldiskfs_xattr_inode_iget: error while reading EA inode -2147483347" Inspection

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/24524/
Subject: LU-8974 osd-ldiskfs: increase supported ldiskfs fs size
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: cfa981f3bf06d602aee998e64d4758e13f48aab8

[Andreas Dilger]

Patch landed to 2.10.

[Gerrit Updater]

Artem Blagodarenko ([e-mail]) uploaded a new patch: https://review.whamcloud.com/36705
Subject: LU-8974 osd-ldiskfs: increase supported size to 1024tb
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: c87bbaf9063494f702c0d0b5884d11dba483e306