[KLUG Members] Redhat 7.2.

[Bryan-TheBS-Smith]

Jim Pharis wrote:
> The peices I managed to extract and remember is that ext3 is not
> SMP ready.

Under kernel 2.4, the Ext2 and Ext3 codebases are the same.  So any
"bugs" in Ext3 also affect Ext2.

Under kernel 2.2, Ext3 is implemented much differently.  Very early
Ext3 kernels (e.g., 0.0.2f) worked flawlessly for me, as they were
little more than a "double buffer".

Probably Ext3's biggest strength is the fact that if the code
detects _any_ problem with the journal, it drops down to a full Ext2
fsck.  I like Ext3's conservative nature in that regard, as well as
the near-instant reversability back to Ext2.

I've already had someone else report they've seen IBM's JFS goto the
journal when it shouldn't have, and nearly toasted the volume
completely (he lost much of his data).  "Aggressive journal
recovery" is the reason I _hate_ NTFS.  I'd much rather wait on an
fsck then lose a volume.

> The other was that with a certain very specific
> system configuration there is a bug in which the hd fills up for
> no appearent reason.

You mean running out of inodes?  Again, Ext2 and Ext3 are the same
codebase for kernel 2.4.

-- TheBS

-- 
Bryan "TheBS" Smith     mailto:b.j.smith@ieee.org    chat:thebs413
Engineer   AbsoluteValue Systems, Inc.   http://www.linux-wlan.org
President    SmithConcepts, Inc.      http://www.SmithConcepts.com
------------------------------------------------------------------
Those living in the US who consider the American flag to be a sym-
bol of oppression obviously fail to understand what the word means