Subject: Re: LFS partition limitations
To: Tracy J. Di Marco White <>
From: Nathan J. Williams <nathanw@MIT.EDU>
List: current-users
Date: 10/03/2000 14:21:24
> I had thought to use FFS+softdep, but my husband thought a journaling
> filesystem would be much safer.  I wanted to use NetBSD on it, and so
> I proposed LFS. compromise.  I'm really not that knowledgeable about
> what a log/journal based file system would gain you.

Journaling vs. soft updates is not settled. See the paper that
was presented at last summer's USENIX technical conference (this paper
was known to some of the Usenix attendees as "filesystem celebrity
deathmatch" due to the "big names" involved):

or if you're a USENIX member:

by Margo I. Seltzer, Gregory R. Ganger, M. Kirk McKusick, Keith
A. Smith, Craig A.N. Soules, and Christopher A. Stein.

From the abstract:

    In this paper, we explore the benefits of Soft Updates and
    journaling, comparing their behavior on both micro-benchmarks and
    workload-based macrobenchmarks. We find that journaling alone is
    not sufficient to "solve" the meta-data update problem. If
    synchronous semantics are required (i.e., meta-data operations are
    durable once the system call returns), then the journaling systems
    cannot realize their full potential. Only when this synchronicity
    requirement is relaxed can journaling systems approach the
    performance of systems like Soft Updates (which also relaxes this
    requirement). Our asynchronous journaling and Soft Updates systems
    perform comparably in most cases. While Soft Updates excels in
    some meta-data intensive microbenchmarks, the macrobenchmark
    results are more ambiguous. In three cases Soft Updates and
    journaling are comparable. In a file intensive news workload,
    journaling prevails, and in a small ISP workload, Soft Updates

        - Nathan