Corruption when transfer Solaris file to Linux 1

[toddyl]

Hi,

I'm trying to gather various information about my servers and have put together a perl script to do this. One of the things I'm interested in is CPU information. Early on I check if the system is Solaris or Linux and run the appropriate commands. On Solaris I am running kstat -m cpu_info. My process runs in such a way that the output is sent to a file on a Linux server:

ssh mySolarisServer /apps/runScript.pl > output

My problem is that when I run this script I am seeing corruption (lots of ^@ symbols) after the text I'm looking for from the Solaris machines.

I logged into my Solaris server and ran:

kstat -m cpu_info > output

The output seemed fine so I then used scp to transfer the file to the Linux box. However when I do this I am seeing the same corruption in the output file on the Linux server.

Does anyone have any idea why this is and what can I do (in Perl) to fix this.

Thanks,

Tom

 

[Annihilannic]

What is in your profile on the Solaris server? Is it enforcing any terminal settings perhaps? Try renaming your .profile (or whichever one is appropriate for your choice of shell) temporarily to disable it on the Solaris server and see if that makes any difference.

Annihilannic.

 

[toddyl]

I've tried what you said (am using a cshrc file) but it made no difference :-(

 

[Annihilannic]

Oh, okay, I just checked and it seems that it's just a 'feature' of the kstat -m cpu_info output. For example, if you do a kstat -m cpu_info | cat -vet on your Solaris box you will see the same characters. They seem to be null characters, so try doing

[tt]ssh hostname 'kstat -m cpu_info | tr -d "\0"'[/tt]

Annihilannic.

 

[toddyl]

Annihilannic,

Your a star. That worked perfectly.

Thanks for your help.

Toddyl