[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
Large disk support in RHEL 3?
- From: Don MacAskill <redhat onethumb com>
- To: taroon-list redhat com
- Subject: Large disk support in RHEL 3?
- Date: Fri, 24 Oct 2003 23:42:07 -0700
Hi all,
I've got some dual-AMD64 boxes hooked up to some big disk arrays, and I
was expected Enterprise 3 to support large (>1TB) block devices on 64bit
platforms.
That doesn't seem to be the case, at least, with the tests I've been doing.
Specifically, I took two ~900GB devices and made a single, ~1.8GB RAID-0
md device. That appeared to work fine. Then, I tried formatting it
with ext3:
[root zeus2 proc]# mkfs -t ext3 /dev/md5
mke2fs 1.32 (09-Nov-2002)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
494616576 inodes, 989209024 blocks
6510778 blocks (0.66%) reserved for the super user
First data block=0
30189 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632,
2654208,
4096000, 7962624, 11239424, 20480000, 23887872, 71663616,
78675968,
102400000, 214990848, 512000000, 550731776, 644972544
mkfs.ext3: Invalid argument while zeroing block 989209008 at end of
filesystem
Writing inode tables: 13805/30189
Could not write 8 blocks in inode table starting at 452362242: Invalid
argument
Next, I tried making it check ever block:
[root zeus2 proc]# mkfs -c -t ext3 -j /dev/md5
mke2fs 1.32 (09-Nov-2002)
mkfs.ext2: bad blocks count - /dev/md5
Finally, I built a custom 2.4.21-4.EL kernel with JFS support (it's in
the kernel tree, don't know why the base kernels don't seem to have it),
and tried JFS:
[root zeus2 root]# mkfs -t jfs /dev/md5
mkfs.jfs version 1.1.2, 25-Mar-2003
Partition must be at least 16 megabytes.
Note that all of the above works fine on those disks individually, and I
have 4 other (smaller) raid devices, so everything seems to be fine
until I go over the 1TB barrier.
I thought I'd see if anyone else has disks large enough to try this on
before I contact support, since I'm not at a mission-critical stage yet.
Anyone?
Thanks,
Don
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]