There's a TSM server that runs on Linux/390? When did this happen? I would think that any write of any kind to any of the journaling file systems would generate an update to the journal, and a corresponding update when the write hits the real disk. There have been some changes in the relative performance of the journaling file systems on Linux/390 lately. Take a look at some of the performance update presentations from IBM to get a feel for which one might be the best choice for you: http://linuxvm.org/Present/SHARE102/s9257jo.pdf
Mark Post -----Original Message----- From: Linux on 390 Port [mailto:[EMAIL PROTECTED] On Behalf Of Thomas Denier Sent: Tuesday, May 04, 2004 2:15 PM To: [EMAIL PROTECTED] Subject: Time stamps and journaling file systems We are probably going to be running a TSM server under mainframe Linux. One of the issues we are looking into is the performance impact of different file systems. TSM manages disk I/O in the same way as many database packages; it preallocates large files and then overwrites specific blocks as needed. As a result there will usually be no update activity for metadata used in managing the allocation of disk blocks. However, the atime, ctime, and mtime fields in the inodes will be updated. Will these updates trigger journaling in file systems that use journaling for metadata changes? If so, how often will these updates happen and how much performance impact will they have? ---------------------------------------------------------------------- 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