[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20040527: bigbird rises from the ashes :-)

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: 20040527: bigbird rises from the ashes :-)
  • Date: Thu, 27 May 2004 10:25:18 -0600
>From:  Gerry Creager N5JXS <address@hidden>
>Organization:  Texas A&M University -- AATLT
>Keywords:  200405270011.i4R0BjtK010667 LDM Linux RAID JFS

Hi Gerry,

>Cleaned upsome fubars in raidtab, rebuilt the array, reformatted, it's 
>now started again.
>
>Let's see how things run.

OK.  Can you tell me the rationale behind leaving the chunk-size in
/etc/raidtab at 4K?  I did see some chatter on Google Groups that had
raidtab listings that used 4K chunk-sizes, but this struck me as not
being a recommendation.  On top of that, the automatic setup of
software RAID by the Fedora Core 1 installation routine resulted in a
64K chunk-size on my dual Athlon 2800+ machine.  When that observation
is coupled with my experience of better performance being seen with
larger and larger block sizes when using the Promise Technology TX2000
RAID card, it just "feels" like the performance would improve by
increasing the chunk-size.  In fact, I was thinking of redoing the RAID
on the machine in my office and increasing the chunk-size to 128K.

I guess that we could let the current configuration run for a few
days and see if the scouring still takes hours.  If it does, I would
like to lobby for increasing the chunk-size and redoing the RAID 
once again (sorry).

>I'll be at 979.229.5301, most all day.  I've got a little writing 
>assignment for the next 2 days, but that's simply gonna make me busy, 
>not inaccessible.  I'll be over near the box, if I need to try anything.

I am at home today waiting for my sewer line to be rooted.  I will be
checking email periodically (not continuosuly) throughout the day.

Cheers,

Tom
--
+-----------------------------------------------------------------------------+
* Tom Yoksas                                             UCAR Unidata Program *
* (303) 497-8642 (last resort)                                  P.O. Box 3000 *
* address@hidden                                   Boulder, CO 80307 *
* Unidata WWW Service                             http://www.unidata.ucar.edu/*
+-----------------------------------------------------------------------------+