This is exactly why journaling file systems were invented.  Make the file
system ext3, reiserfs, xfs, or jfs (did I miss any?), and startup time after
a system crash should be negligible.


Mark Post

-----Original Message-----
From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Thomas
Denier
Sent: Monday, October 11, 2004 4:30 PM
To: [EMAIL PROTECTED]
Subject: Performance of ext3 crash recovery


We are considering setting up a mainframe Linux system with a file system
with a size somewhere in the hundreds of gigabytes. This file system would
contain a few hundred files with sizes ranging from a couple of hundred
megabytes to several gigabytes. Some of our staff have expressed concerns
about the elapsed time required to mount such a file system if we had to
reboot after a crash, requiring the journaling mechanism to clean up
inconsistencies in the contents of different disk blocks. Is there any data
available on the elapsed time for this clean up process, and its dependency
on different workload characteristics (such as number of blocks, number of
files, and number of open files)?

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions, send email
to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to