user28362's questions

Is there a way to limit the speed of scp (or rsync)? I couldn't find any information in the man page, nor on the internet.

I am using scp for backuping data. When I get the cron output(using crontab file for cron operations) with mail, I do not get any output. The scp operation is done succesfully.

So, in the terminal I get for example: test.file 100% 1489KB 1.2MB/s 00:00

But in the cron output, I do not get the output of the command.

Does someone know how to do this or which other way do you use to get information about the command?

Thanks.

I would like to know if it possible to let a Windows Xp machine authenticate to Squid (Linux) using Kerberos without the need of an Active Directory domain.

I only want to create a Kerberos ticket on the client side, which should give the client access to squid (using I.E.).

I only found tutorials about configuring A.D./Squid, not an environment with only Linux servers.

Thanks

Update:

The kerberos setup is correctly done, the proxy and client can get tickets.

As for the browser (FF/IE), I get:

ERROR
Cache Access Denied

While trying to retrieve the URL: http://www.google.com/

The following error was encountered:

    * Cache Access Denied. 

Sorry, you are not currently allowed to request:

    http://www.google.com/

from this cache until you have authenticated yourself. 

In kerberos, I get:

squid_kerb_auth: Got 'YR ElRNTVMTUABBAABAB4IIogAAAAAAAAAAAAAAAAAAAAAFASgDAAAADw==' from squid (length: 59).
squid_kerb_auth: parseNegTokenInit failed with rc=101
squid_kerb_auth: received type 1 NTLM token

This message is strange, as I didn't configure NTLM. It looks like the browser uses the wrong authentication methode.

I have a folder with one big file. This is a file that changes each day. Using tar for incremental backups is not a good idea as that big file (each file that is changed is saved) will be copied each day that an incremental backup is made.

Is there an application that makes a diff of two binary files (a differential backup)?

Thanks.

I will configure a bond of three NIC's. I want to define which one will be first used and which NIC will be used as the second and third NIC when failover is done. Is this possible?

Thanks

I use two Q-Logic HBA's on one server which have access to the same LUN. Red Hat (and gparted also) sees the LUN as two seperate disks (sda, sdb). Is this normal?

Thanks!

I want to deploy an Active Directory. There are some users of an other A.D. domain that should get access to the computers that will be part of my own A.D. domain. Can I configure this?

Thanks,

I am experimenting with LDAP. As I see it, apache uses the Unix password field to get the credentials and samba the field that is specific for samba.

I just use the LDAP server for a single server, LDAP runs in the same server. What I wanted is to use the same password for apache and samba (and mediawiki etc). Is it possible? If not, are there any alternatives?

Thanks!

I would like to disable the "gdmgreeting" application that is a graphical frontend, as I only want console interfaces. This application runs on vt7 and restarts every time after I manually kill it. I have no desktop environment installed (gnome, kde), so this app just uses memory and CPU for nothing.

What can I do? I couldn't find any initscript or something like that that runs it.

The following processes are causing it to appear: /usr/bin/Xorg : 0 -br -audit 0 -auth /var/gdm/:0.Xauth -nolisten tcp vt7

/usr/libexec/gdmgreeter

Removing the full graphical frontend is a better solution (x11?).

Thanks!

I will deploy a web server, but I don't have much experience.

I would like to know what happens with mail that is send when the server is offline (restarting, maintenance, etc.).

When an internal user wants to send and receive mail, it's logic that he can't access the server. But when an external person uses his mail server to send a mail to my mail server, will the mail be resent on a later time?

The mail server will be a virtual machine, is this a good idea?

Thanks.

I have a server with 4 ports (2 adapters with two ports each). I want to connect two ports to the local switch and two other ports to the external switch. This local switch has an uplink to the external switch, which has access to the net. As the local switch has a faster connection (1 Gb), I want to use it as primary switch using the two ports. If this switch fails, I want that the following two ports, that are connected with the external switch, will get activated.

Is this possible?

Thanks.

I use Squid and the users must authenticate. The users are authenticating in cleartext. I found tutorials about stunnel (http://www.jeffyestrumskas.com/?page_id=3), but the clients also need this software.

Is there no possibility that for user authentication SSL (or some other encryption technology) can be used?.

Thanks.

I would like to know if it is necessary to align LVM's in it?

I'm using VMware vSphere and as it is a best practise to align the disks (to 64 or 128 kb) I want to align every disk.

Thanks

EDIT: I'm using VMware vSphere and as it is a best practise to align the virtual disks (to 64 or 128 kb).

I would like to know if it is possible to use the FC-switch (that is used for a SAN) also for TCP/IP traffic.

We have servers (virtual machines) that must have a fast network for copying much data. It would be better to use the existing FC network than a slower 1 Gb network.

Is this possible?

For testing purpose of an application, I would like to limit the RAM of a server. I prefere not to remove the RAM-sticks. I could not find an option in the BIOS. Is there a way to modify this in the Linux configuration?

Thanks.

I searched the internet, but I couldn't find which sort of cables iSCSI adapters require. I will use it in a HP server, which has two HP NC382i Integrated Dual Port PCI Express Gigabit Server Adapters. Does it use CAT 5 or some special cabling and port (UTP)?

Thanks very much!