skinp's questions

I have a mounted NTFS partition under Windows (2003) and Linux (RHEL 5.7) at the same time. The LUN resides on SAN and is presented to both servers at the same time. At first, the setup looks ok as I initially can see files from both Windows and Linux.

Writes are only initiated from Windows and I mounted the partition under Linux with read only (ro) and no access time (noatime) options. But when I add files from Windows, I can't see them under Linux. Is there anything I'm missing? Is it even possible at all?

I would like to achieve this in order to avoid copying from the network for a data migration of > 1Tb. Do you see any other way to do it?

We have a spare Integrity blade (2x Tukwila quad-core + 16Gb RAM) laying around which I would like to use has a Virtual Host for a couple of Linux VMs. I am having some concerns finding the best solutions for our situation (if a solution is possible at all?). Here's what I'm dealing with:

• XEN looks well supported on Itanium, but I'm not sure if I would be able to run x86_64 guests on ia64+XEN.
• I guess KVM would be able to support emulation for x86_64 through qemu, but support for ia64 does not seem to be well supported.
• All guests are going to be RedHat 5.5 based and HAVE to be on x86_64 architecture.
• I don't care about the host's distribution (ideally CentOS or Debian), but I would prefer it not being Red Hat (since I don't want to deal with the trouble of registering only for 1 ia64 licence).
• Red Hat drops support for ia64 and Xen in version 6
• Debian seems well supported on ia64 but I can't seem to find kvm or xen packages for it
• Will CentOS also drop support for ia64 and Xen?

So, to sum up my question, how would you guys go about virtualising x86_64 guests on an Itanium server?

When adding a new drive, it took over /dev/sda1. Now my old drive is /dev/sda2 and it messes with the current fstab setup of the server. I know I can get away with using UUID or labels (what I just did), but I would still prefer having the primary drive take /dev/sda1...

So, how do I force a disk to take a particular device number?

I'm looking to transfer data across 2 lv of an HP-UX server. I have a couple of those transfers to do, some of which are mostly binary (Oracle tablespace...) and some others are more text files (logs...). Used data size of the volumes is between 100Gb and 1Tb. Also, I will be changing the block size from 1K to 8K on some of these partitions...

Things I'm looking for:

• Guarantees data integrity
• Fastest data transfer speed
• Keeps file ownership and permissions

Right now, I've thought about dd, cp and rsync, but I'm not sure on the best one to use and the best way to use them...

I want to delete the top level directory which does not contain any files, but can contain other empty (meaning again not containing files) directories. For example:

$ ls -R
.: foo/
./foo/: bar  bar1
./foo/bar/:
./foo/bar1/:

Here, I would like to delete the directory foo/ (and its subdirectories). At first, I thought of using something like find . -type d -empty for the search, but since foo contains directories, it only finds the lower level ones:

$ find . -type d -empty
./foo/bar
./foo/bar1

I guess I could loop until find . -type d -empty finds nothing, but I may end up having a very big directory structure containing a lot of those empty directories and I'm concerned about the performance impact of doing it that way...

Any idea?

I want to know if a HP-UX machine can communicate to a remote machine (which I don't have control over) on a particular port using UDP.

I tried telnet, but it doesn't look like it supports UDP. I would use netcat, but I was unable to find any bundle. This machine is in production so I don't want to mess around compliling or installing things too much... (nmap, netcat from source...)

Any simple way to do this?