Adam Thornton wrote:
> This is mentioned in a WebSphere context.  Is there anything in
> WebSphere so rude as to bypass the filesystem and try to write directly
> to the disk device?  Or perhaps WebSphere's database attempting to use
> the device in raw mode?  In short, is there any failure mode other than
> user error that might directly write to the first few records on the
> disk?  zipl obviously *does* write there, but it also preserves the
> existing format type just fine.

While I'm not familiar with the WebSphere installation process, I find it
highly unlikely that any user space program, when given a filesystem,
will willingly try to move up to the partition, the device and
arbitrarily overwrite data found there. Any such program would need to
include highly platform specific code which would make it difficult to
maintain across platforms. Even databases that access a device in raw
mode need to be explicitly given the device file to use instead of
deriving it from the filesystem.

zipl is different as it is part of the s390-tools package which is
specifically released for a particular version of the operating
system.

We'll try to contact the author of that WebSphere installation problem
report to verify our assumption. I think it is safe to assume that the
only way this problem occurs is following a user error.


Regards,
  Peter Oberparleiter

--
Peter Oberparleiter
Linux on zSeries Development
IBM Development Lab, Boeblingen/Germany

----------------------------------------------------------------------
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