Re: URGENT!! Synaptic Error : Subprocess - /bin/rpm Error 08
On Sat, 18 Oct 2003 21:29:10 -0700 Shesh Kondi [EMAIL PROTECTED] wrote: Hi, My synaptic (apt) was working fine till yesterday. Now, I am getting this error message when I am trying to upgrade packages on my RH 9 laptop. Sub-process /bin/rpm returned an error code (08). Well i guess if it's urgent!! you could try changing the following line in /etc/apt/apt.conf to false: GPG-Check true; Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: sending mail with attachment via commandline?
On Thu, 16 Oct 2003 11:42:22 -0700 Chris W. Parker [EMAIL PROTECTED] wrote: Hey everyone. I've got a shell script that does a simple backup of a directory on a nightly basis and since I don't have samba configured I'd like to send that backup file to my email address. I've 'man mail' but didn't see anything about adding attachments? Did I miss it or do I need to use a different command? I'm only familiar with 'mail' and being a command that can send mail from the command line. Hi Chris, You can use mutt : mutt -a $ATTACHFILE -s $SUBJECT $MAILTO $TXTFILE Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: sending mail with attachment via commandline?
On Thu, 16 Oct 2003 12:05:36 -0700 Chris W. Parker [EMAIL PROTECTED] wrote: I already have pine installed. Will there be any sort of conflict if I also install mutt? Chris, It's hard to imagine any problem, especially if you're just using Mutt to _send_ the occasional message. Don't think it's any different than having Pine and Mailx installed on your system at the same time. HTH, Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Firewall - Limit Geographic Area
On Wed, 15 Oct 2003 15:47:56 -0500 lrnobs [EMAIL PROTECTED] wrote: Does anyone know of a way to do this? Are the IP ranges assigned to American networks published somewhere? Yes, They are published in many places but one detailed list is available freely from: http://ip-to-country.directi.com/ This is in a machine readable format with updates. According to this list there are over 6600 ranges of IP's assigned in the USA and about 170 in China for instance. It's pretty easy to automate the translation of this database into iptable rules. As others have already mentioned, you'll have to think carefully about how much benefit this really provides and the problems it might cause. The database is more useful to customize web content based on the location of a visitor. It might be better to use the data to redirect foreign visitors to a different page on your site which explains why they probably won't find anything of interest. Regards, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: passwdqc problem
On Wed, 15 Oct 2003 23:41:45 +0200 Janus N. Tøndering [EMAIL PROTECTED] wrote: Hi, I've been trying to install the passwdqc module from Solar Designer and configure it with pam. I have big troubles though: it does not update the password to the newly entered one. Nothing shows up in logfiles. $ cat /etc/pam.d/passwd auth required pam_stack.so service=system-auth accountrequired pam_stack.so service=system-auth password required /lib/security/pam_passwdqc.so ask_oldauthtok=update check_oldauthtok Hi Janus, Try using the following lines along with the one you showed. pam_passwdqc does not update the password itself but relies on pam_unix to do so: password required pam_passwdqc.so ask_oldauthtok=update check_oldauthtok password sufficient pam_unix.so use_first_pass md5 shadow password required pam_deny.so Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: ip alias +routing
On 16 Oct 2003 02:04:32 +0200 Chema Carballido [EMAIL PROTECTED] wrote: Hello, I would like to route two networks using one network -card. I think i can set to diferent ip address for that card using alias. But how can i enroute the traficc from one network to another. Hi Chema, All you should need is to enable routing between configured interfaces: echo 1 /proc/sys/net/ipv4/ip_forward or sysctl -w net/ipv4/ip_forward=1 Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: bash reverse menu-complete
On Wed, 01 Oct 2003 20:04:15 -0300 Herculano de Lima Einloft Neto [EMAIL PROTECTED] wrote: Hello, I would like to bind menu-complete with a -1 argument to something, S-TAB for instance. Is there any way to do that? Other shells offer reverse-menu-complete. If I could bind readline arguments.. bind S-TAB:'menu-complete(-1)' - is there something like this? Hi Herculano, This is a bit of a tough one, and depends on how you have readline configured (ie. emacs or vi mode). If you add the following to the end of your /etc/inputrc file you should get what you want: $if mode=vi \C-0-: digit-argument TAB: menu-complete \e[Z: \C-0-\t $else TAB: menu-complete \e[Z: \M--1\t $endif man readline has all the details about why this works. HTH, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Why does rhnupdate create sendmail.cf.rpmnew instead of sendmail.mc.rpmnew? How to handle.
On Wed, 01 Oct 2003 17:20:48 -0700 Mike Klein [EMAIL PROTECTED] wrote: Given that sendmail.cf is by most sane individuals generated via the file sendmail.mc, what is the impact of updating sendmail via rhnupdate when I only get a sendmail.cf.rpmnew instead of a sendmail.mc.rpmnew? Seems like I might be missing some crucial sendmail parms for security/etc. I certainly don't want to be backporting cf changes into an mc file. My sendmail book is like 3' thick...and I'm not worthy. Hi Mike, rhn up2date is based on rpm package management. rpm is pretty smart about upgrading your system. It knows what files are configuration files that have been modified by you. In order to preserve your changes it does not overwrite configuration files that you've changed. It will add the .rpmnew extension to the new file or save your current file with .rpmsave. Either way it will ensure you don't lose work, but you may have to do some tweaking after the upgrade. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: modules-init-tools
On Thu, 2 Oct 2003 09:48:53 -0500 Paillet, Daniel [EMAIL PROTECTED] wrote: I am trying to compile 2.6.0-test5 kernel on red had 9. I was told that I needed to have modules-init-tools to compile the kernel successfully. I have downloaded the modules-init-tools. Where to I install it ? ? Daniel, You can use the modutils rpm package from rawhide and install it like any other rpm. It is only one version behind the current M.I.T. and has some specific RedHat patches: ftp://ftp.redhat.com/pub/redhat/linux/rawhide/i386/RedHat/RPMS/modutils* Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Why does rhnupdate create sendmail.cf.rpmnew instead of sendmail.mc.rpmnew? How to handle.
On 02 Oct 2003 10:43:27 -0500 Bret Hughes [EMAIL PROTECTED] wrote: On Thu, 2003-10-02 at 10:17, Sean Estabrooks wrote: On Wed, 01 Oct 2003 17:20:48 -0700 Mike Klein [EMAIL PROTECTED] wrote: Given that sendmail.cf is by most sane individuals generated via the file sendmail.mc, what is the impact of updating sendmail via rhnupdate when I only get a sendmail.cf.rpmnew instead of a sendmail.mc.rpmnew? Seems like I might be missing some crucial sendmail parms for security/etc. I certainly don't want to be backporting cf changes into an mc file. My sendmail book is like 3' thick...and I'm not worthy. Hi Mike, rhn up2date is based on rpm package management. rpm is pretty smart about upgrading your system. It knows what files are configuration files that have been modified by you. In order to preserve your changes it does not overwrite configuration files that you've changed. It will add the .rpmnew extension to the new file or save your current file with .rpmsave. Either way it will ensure you don't lose work, but you may have to do some tweaking after the upgrade. Sean I think you missed his point. perhaps a better subject would be why did I not get a sendmail.mc.rpmnew or why did sendmail upgrade overwrite my sendmail.mc. If this happened then I suggest a bug report be filed. Bret I didn't miss it i ignored it... here's one possibility: no change to senmail.mc make -C /etc/mail (regenerates sendmail.cf with new time stamp) rpm upgrade Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: How to add persistent static routes?
On Wed, 1 Oct 2003 08:44:59 -0400 Ken Morley [EMAIL PROTECTED] wrote: I'm using RH7.3. There are multiple paths off our network. Besides the default gateway, which points to the internet, we also have a router that connects the company WAN to branch offices. I can use the route statement to setup a static route as follows: route add -net 10.184.0.0 netmask 255.255.0.0 gw 10.0.0.254 metric 1 However, this route doesn't stick when I reboot the system. How do I make this route persistent so it will always be there (even after a reboot)? Hi Ken, You setup the static routes you want in /etc/sysconfig/static-routes one route per line. For those routes not associated with a transient interface you prefix the route with any so that it is brought up at boot time. All the options of the route command are available, but omit the leading hyphen in front of -net or -host. an example /etc/sysconfig/static-routes : any net 10.10.10.0 netmask 255.0.0.0 gw 192.168.1.1 Hope this helps, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: question on symbolic link
On Sat, 27 Sep 2003 19:28:58 -0400 TK [EMAIL PROTECTED] wrote: Let's say file A1 and A2 are symbolic links to file R. Is there a way to tell which files are linked to R by examing file R only (stat doesn't seem to reveal anything, but I noticed hard link will increase the Links number), instead of searching through all the files in the FS? Only way to find them is to search. Especially if the FS containing A2 is not mounted yet. This is one case where you can't use a hard link and symbolic links are useful. Or symbolic link is a one-way knowledge that only A2 knows about it but R has no clue at all? Correct, the symbolic link holds a pointer (ie. the path) to the target. The target knows nothing about any symbolic links referencing it. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Syslogd UDP Port
On Thu, 25 Sep 2003 16:19:18 -0500 Brett Franck [EMAIL PROTECTED] wrote: If it is possible, how would you map say UDP port 515 to syslog and have syslog report it to a specific logfile through the syslog.conf file? Hi Brett, Here's one way to do it using the standard syslogd: Attached is a script file that will build a chroot environment for the syslogd command. Run this build script to create a subdirectory with everything you need to run another copy of syslogd on a separate port. # ./build syslog2 CD into the syslog2 subdirectory and run the ./go script to start a syslogd daemon listening on port (port 515 is reserved for printer, check your /etc/services file): # cd syslog2 # ./go You can create and run as many copies of this chroot environment as you like. You'll find a log file under each directory you create as ./var/log/all.log. You can tweak the files under ./etc to configure the logging and port no. I just whipped this together over a coffee this morning so no guarantees at all, but it looks simple enough that it should run for you. If you hit any snags, happy to help. Good Luck, Sean P.S. The build script will create the following directory structure, and copy/create proper files: syslog2 |-- dev |-- etc | |-- nsswitch.conf | |-- services | `-- syslog.conf |-- go |-- lib | |-- ld-linux.so.2 | |-- libnss_files.so.2 | `-- tls | `-- libc.so.6 |-- sbin | `-- syslogd `-- var |-- log `-- run build Description: Binary data
Re: Nautilus CD recording?
On Fri, 26 Sep 2003 16:19:01 + Nick Lindsell [EMAIL PROTECTED] wrote: On Fri, 2003-09-26 at 16:05, Nick Lindsell wrote: Greetings all, I was pleasantly surprised that Nautilus file manager now has CD writing capabilities. But mine complains there is not enough space to create the ISO image - would someone enlighten me on where Nuatilus is trying to create the image so I can ensure there is enough room. Perhaps it is an environment variable I seek? Found it - it writes image.iso in /tmp. Which is dumb, imho. So can anyone tell me how to change that? One way: # export TMPDIR=/where/ever/you/want # nautilus-cd-burner Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: rpm problems
On Fri, 26 Sep 2003 09:21:54 -0600 Aly Dharshi [EMAIL PROTECTED] wrote: Hi Folks, I have this problem as mentioned: rpmdb: unable to join the environment error: db4 error(11) from dbenv-open: Resource temporarily unavailable error: cannot open Packages index using db3 - Resource temporarily unavailable (11) error: cannot open Packages database in /var/lib/rpm warning: up2date-3.1.23.2-1.i386.rpm: V3 DSA signature: NOKEY, key ID db42a60e I tried to kill all the __db files and do a rpm -rebuilddb but that produces the error: rpmdb: write: 0xbfffb8a0, 8192: Invalid argument error: db4 error(22) from dbenv-open: Invalid argument error: cannot open Packages index Run this before you use rpm : export LD_ASSUME_KERNEL=2.2.5 Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Chroot Jail Time - extra syslogd instances
On Fri, 26 Sep 2003 14:11:44 -0500 Brett Franck [EMAIL PROTECTED] wrote: How do you teach a chroot jail the correct time? Brett Hey Brett, I know your question is in regard to the syslogd chroot environment we talked about this morning. Attached is an updated build script that adds support for your timezone and should be slightly more portable than the first version. It addresses the time problem by setting the TZ variable and copying the proper zoneinfo data into the chroot jail. It will default to the Eastern timezone unless you override it with a 3rd command line parameter, for example: ./build syslog3 US/Central Good luck, Sean build Description: Binary data
Re: help in kernel compiling
On Thu, 25 Sep 2003 19:58:11 +0530 Nabin Limbu [EMAIL PROTECTED] wrote: Hi all, I have upgraded my kernel to 2.4.20-20.9 via up2date with no problem in RH 9. In this new kernel, I wanted to enable a cyclades card in kernel for which I complied this kernel as below:- - make menuconfig - enabled Cyclades card in the menu - make dep; make clean - make bzImage - make modules - make modules_install - cp /usr/src/linux.x.x./arch/i386/boot/bzimage /boot/ - Edit /etc/grub.conf - Replaced /boot/vmlinuz-2.4.20-20.9 with /boot/bzImage - Rebooted While trying to boot with this new kernel, I got following message: root (hd0,0) Filesystem type is ext2fs, partition type 0x83 kernel /boot/vmlinuz-2.4.20-20.9 ro root=LABEL=/ Error 13: Invalid or unsupported executable format Press any key to continue What might be the problem? How can I solve it? Please help. Hi Nabin, Make sure compiled in : Kernel support for ELF binaries under the General Setup option of menuconfig. which will add the following line to your .config: CONFIG_BINFMT_ELF=y That might be all that's wrong, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: group permission and directory
On Thu, 25 Sep 2003 11:30:58 -0400 Reuben D. Budiardja [EMAIL PROTECTED] wrote: Hello, I have the following problem which I hope someone can help. Suppose I have users: a,b,c,d,...,z. User a,b,c,d,e,f,g is member of group cooluser The other users are member of the default group (which in RH is the group with the same name of the username). I want user d,e,f,g,h,i,j,k,l only to be able to share stuff in a directory, so I create a new group: sausage (using groupadd command), and put those user as a member of the group sausage. I created directory /home/sausage and change the owner to root.sausage with complete access to the group $ chown -R root.sausage /home/sausage $ chmod -R 775 /home/sausage Now, the problem is, if user d put a file in the /home/sausage, the ownership of the file is d.cooluser instead of d.sausage, [EMAIL PROTECTED] sausage]# ls -lah total 16k drwxrwxr-x2 root sausage 4.0k Sep 25 11:01 . drwxr-xr-x 46 root root 4.0k Sep 25 10:59 .. -rw-r--r--1 d cooluser6.6k Sep 25 11:01 KlugGlosT.DOC thus the other member of sausage which are not member of cooluser can't write to the file. What do I need to do so that the default ownership and permission in the directory sausage is root.sausage and rwx for the group? Thanks in advance for any help RDB -- Reuben, You need to tell Linux that all files created in that directory should be created with the Directory group instead of the users group: chmod g+s /home/sausage Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: group permission and directory
On Thu, 25 Sep 2003 12:04:41 -0400 Reuben D. Budiardja [EMAIL PROTECTED] wrote: Thanks, that's work. But only the group permission is still r instead of rw. I read man chmod and play around but still got nowhere. Do I need to change the umask for the user instead? Yes, you'll want an appropriate umask for the users, i'm surprised that it isn't correct already. umask 002 This should be set already if the userid in question is above 100 and you're not using su to test. Also, if I do ls -la /home, I see the sausage directory listed as drwxrwsr-x2 root sausage 4.0k Sep 25 12:10 sausage my question is, what's s in the group permission? On a directory that simply means that the directory group will be used for all files created in this directory. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Kernel Update - if it's not broke do I fix it?
On Thu, 25 Sep 2003 12:18:46 -0400 Billy [EMAIL PROTECTED] wrote: I have a 7.3 server running Apache, PHP, MySQL, and WU-FTP. I of course keep all of those packages updated since I have to have most of the ports open in the firewall to use them. However, I have not upgrade the kernel since the install about a year ago. It is version 2.4.18-3. I have three questions, #1 How important is it to keep the kernel updated, I could imagine maybe I have just gotten lucky but the machine has been wonderful up to date. #2 How big of risk do I run upgrading the kernel, is there a good chance that I could hose the machine, or is it as easy as running rpm-Fvh kernel.rpm? #3 When I run rpm -qa | grep kernel it only returns kernel-2.4.18-3, when I look at the available kernel updates I find kernel-BOOT-2.4.20-20.7.i386.rpm and kernel-source-2.4.20-20.7.i386.rpm as well...my practice in the past has always been to only install updated packages for what I already have on the machine...excuse my ignorance but I am extremely new to the linux world, and have heard so many horror stories about kernel updates...any input would be greatly appreciated!! Hey Bill, It is quite probable that there have been security related updates to the kernel that would be relevant to your environment. You have to decide for yourself how important that makes it for you to upgrade. Obviously there is no other reason to upgrade as you're content with the way the system is operating. If you do choose to upgrade your kernel i'd suggest _adding_ the new kernel but not removing your existing kernel: rpm -ivh kernel_new_zzz.rpm This will give you a way to retreat if there are any problems with the new kernel. You _dont_ want to use the BOOT kernel, that's pretty much just for installation media like CD's. And you probably don't have any reason to install the kernel-source package. HTH, Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Kernel Update - if it's not broke do I fix it?
On Thu, 25 Sep 2003 14:38:10 -0400 Jason Murray [EMAIL PROTECTED] wrote: I have the same problem, but my further questions are: will installing the mew kernel with rpm -ivh kernel_new_zzz.rpm (a) add the new kernel image to my boot loader (in this case GRUB) (b) keep the entry for the old kernel in my boot loader I don't know of any other way to be able to retreat if for some reason the new kernel does not boot... Hi Jason Yes to both questions. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Kernel Update - if it's not broke do I fix it?
On Thu, 25 Sep 2003 15:36:58 -0400 Billy [EMAIL PROTECTED] wrote: Thanks for the advice Sean...I am going to research the kernel updates some more and evaluate the upgrade. I do have a question about the kernel and being able to retreat. If I install the newkernel.rpm with rpm -ivh it will install, and running in conjunction with the old kernel? So once I reboot by default it will load the newest kernel? Then once I let it run for a couple days and everything seems fine I could safely run rpm -e oldkernel.rpm to remove the old? Then on the other hand, if I install the newkernel.rpm and something isn't working right could reboot into the old kernel and run rpm -e newkernel.rpm? Yes, it will work as you describe. When the grub menu is shown to you at boot up time you'll be able to select which kernel to use but the new one should be the default if you select nothing. Is there a chance that after the kernel has been updated that the machine will not boot at all, or as long as I have the old kernel I can always boot with that? Installing a new kernel _should_ do nothing to stop you from rebooting and using the previous kernel. Still that's no excuse not to have proper backups ;o) And finally *if* I go ahead with this am I crazy to do this remotely over SSH? If for some reason the new kernel hangs on reboot, you'll have to be in front of the console to reset it. I've upgraded kernels remotely for years, but on occasion have had to get in the car and go to the server or call someone for a reset. Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Power-off on shutdown -h
On Thu, 25 Sep 2003 18:52:05 +1000 (EST) Michael Mansour [EMAIL PROTECTED] wrote: From a linux shutdown -h now, they go to the Power down prompt after shutting down the drives, but they do not power-off. See what happens when you use the poweroff command instead of shutdown. And if that doesn't work, try a poweroff -f. The -f will force the issue and prove if the problem is with your init scripts or something at a lower level. But be a bit careful with the -f because things don't get shut down gracefully, you wouldn't want to use it regularly, only to diagnose where the problem is. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Problems with CardBus NIC
On Wed, 24 Sep 2003 08:42:20 -0400 David Hart [EMAIL PROTECTED] wrote: We're getting quite a few of these on one client. This works perfectly in Windows. Card is a 3Com 3CXFE575BT. It seems to be identified properly. Sep 24 07:41:58 main kernel: eth0: Interrupt posted but not delivered -- IRQ blocked by another device? Sep 24 07:41:58 main kernel: Flags; bus-master 1, dirty 236702(14) current 236702(14) Sep 24 07:41:58 main kernel: Transmit list vs. c02d05 There should have been two additional error lines above the one you started with. The first one saying transmit timed out. They would give important information but my guess would be tthat you have some bus-mastering device in the machine that is causing problems. This seems like a hardware configuration problem (time to get your manuals out) but there's also a chance you've hit a kernel bug. Make sure you've updated to the latest kernel (because that's an easy first step) but if the problems continue you'll have to look at how the machines are configured.You may have to swap PCI card locations or assign fixed IRQ's for instance. One BIOS option to pay attention to is PCI 2.2 enable/disable if it's available. Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Problems with CardBus NIC
On Wed, 24 Sep 2003 10:45:55 -0400 David Hart [EMAIL PROTECTED] wrote: Sep 24 09:19:30 main kernel: eth0: transmit timed out, tx_status 00 status e681. Sep 24 09:19:30 main kernel: diagnostics: net 0cc2 media a800 dma 003a. Sep 24 09:19:30 main kernel: eth0: Interrupt posted but not delivered -- IRQ blocked by another device? Sep 24 09:19:30 main kernel: Flags; bus-master 1, dirty 19656(8) current 19656(8) Kernel = 2.4.22 using the latest source and modules from RH Rawhide. The problem has been consistent regardless of which kernel is used. I switched pockets - no difference except I see that hotplugs work just fine ;-) What's frustrating is that the system doesn't specify which IRQ wasn't delivered. Needless to say this is on my client which is always the oldest machine in the place. When I have some time, I'll tweak config.opts. I see that it has the extended port range required for Token Ring for starters. You can try reloading the module with the watchdog parameter to see if it removes the problem for you: modprobe 3x59x watchdog=1 Obviously you have to unload the module before you can reload it. The watchdog option may or may not help but the results would be interesting. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Machine Name resolution
On Wed, 24 Sep 2003 09:09:58 -0700 James D. Parra [EMAIL PROTECTED] wrote: Hello, We're having trouble getting machine name resolution from an DNS server. If we try to ping, for example, a machine named Mach1, we get an unknown host Mach1 error. If we ping the FQDN, Mach1.domainname.com, we get resolution. Oddly, windows boxes can resolve the machine name, from a dos prompt, without the domain suffix (No WINS being used either). What can we do to get machine name resolution, internally, without the domain name from our Linux boxes? Thank you in advance. James Add the line: search domainname.com to the /etc/resolv.conf file Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Scripting questions
On Wed, 24 Sep 2003 17:18:36 +
Abdussamad Abdurrazzaq [EMAIL PROTECTED] wrote:
Hello
Please help me with these basic scripting questions. How do you tell the
path of the current directory?
Assuming you mean Bash scripting
# echo $PWD
Is there an equivalent to the Left$ Right$
functions found in BASIC in Linux scripting? What I mean is how do you
extract a given number of characters from the left or right of a string?
# a=ABCDEF
# echo ${a::2}
AB
# echo ${a:${#a}-2}
EF
Cheers,
Sean
--
redhat-list mailing list
unsubscribe mailto:[EMAIL PROTECTED]
https://www.redhat.com/mailman/listinfo/redhat-list
Re: modem help
On Wed, 24 Sep 2003 10:22:52 -0700 (PDT) Srinivas S [EMAIL PROTECTED] wrote: I have Redhat 9.0 and a Dlink external 56K data/fax modem. (DFM - 560ES) I want to know how i can communicate with the modem either as shell commands or as a C program. I tried the following, which did not work ln -s /dev/ttyS1 /dev/modem echo ATX1DT6565206 /dev/modem is there any other way. Please let me know. Hi Srinivas, You can install the uucp command and then: cu -l /dev/ttyS1 dir will give you an interactive connection to the modem where you can type commands and see results. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: yum/apt-get (was Re: Fedora)
On Wed, 24 Sep 2003 13:37:42 -0400 Buck [EMAIL PROTECTED] wrote: Up until Monday Up2Date was free for filling out a questionaire every two months. The $60 provided you with convenience and earlier access to binary downloads and free binary downloads of RHEL. Up2Date Demo was free for 2 months for each installation with a unique email address. At the end of 2 months, one would fill out a questionaire to extend the subscription for two more months. That was discontinued this week. As Alan said, its a good thing to support Red Hat anyway. Hopefully Fedora will pick it up from there. To me, it would make sense that Fedora picks up the up2date program if for no other reason than to attract financial support from those of us willing and able to pay the $50 - 60 per year. That's my $.02 worth Hey Buck, You'll get _at least_ the functional equivalent of up2date with Fedora. apt-get and yum access is already part of the Fedora Project. But RedHat will not be maintaining updates as long as they were for previous offerings or making any SLA commitments. If this is good enough for your needs, you'll save the money you were previously spending each year. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 15:25:15 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: No, Benjamin has a point: another change (d) is that the free version will not be branded Red Hat, as I also mentioned earlier. This _could_ believably slow down adoption due to brand recognition, both in the case of newbies choosing the well-known Red Hat for their Linux, but also of people like Benjamin's boss (and mine, by the way) who choose or would choose RHEL after seeing RHL at work, but who will not see Fedora in the same light. I think the answers you gave to people later in the day suggest that this isn't anything to worry too much about: On Tue, 23 Sep 2003 00:22:05 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: Why? Personally, I see Fedora as being functionally equivalent to RHL 10, and for small servers I would have run 10 without question. No reason on Earth for me not to use Fedora for those, since I expect to see RH put as much into Fedora as they did into RHL... so no loss in functionality or reliability or trustworthiness. What do you see differently? On Tue, 23 Sep 2003 01:08:55 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: I don't expect them to do that, since this would alienate a huge part of their community which in some way does generate leads for later RHEL sales. I don't think either of us will convince the other, so let's agree to disagree, realize that at least we understood each other, and wait to see what reality turns out to be. Regards, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: username with CAPS
On Mon, 22 Sep 2003 21:25:27 -0500 Bob Hartung [EMAIL PROTECTED] wrote: Hi all, Is there any way to get around adduser's dislike of a capital letter in a user's name? I don't see the solution in the man page. Hi Bob, I suspect you mean the command useradd. It will always enforce the user naming conventions, which basically restricts you to lowercase letters, numbers, hyphen, and underscore characters. If you circumvent these restrictions other tools that make assumptions about usernames may fail. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: gdm_config_parse: The gdm user should not be root
On Tue, 23 Sep 2003 00:01:20 -0800 Noah [EMAIL PROTECTED] wrote: running redhat 8.0 been looking around the web to find a cure and some of the responses dont completely match the files on my machine so I am at a bit of a loss as to what to do here. snip I copied the /etc/X11/gdm/factory-gdm.conf to /etc/X11/gdm/gdm.conf and that did not help so far. what else should I do? Hi Noah I don't have a RH8 box handy but you should see something like this in your /etc/X11/gdm/gdm.conf file: # User and group that gdm should run as. Probably should be gdm and gdm # and you should create these user and group. Anyone found running this # as someone too privilaged will get a kick in the ass. This should have # access to only the gdm directories and files. User=gdm Group=gdm Make sure that these users do exist on your system and have a non zero user id. Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 09:07:33 -0400 Johnathan Bailes [EMAIL PROTECTED] wrote: I don't know I tend to agree with the original sentiment. In comparison to both linux distros and the competition from closed sources the RHE price tag seems a bit high and at the same time fedora might seem a bit too bleeding edge. You are going to have some big markets for linux in the small to medium size deployment areas running around and trying out other choices. This is bad. We use Suse at work here and the professional version is like about $70 bucks. The question will be is the extra $100+ worth it? Most firms are not going to look at the fedora choice seeing it as a bleeding edge desktop style distro. Not what they want on their web servers. Yes, the turn over in terms of boxed units is too high and there needs to be another choice. However, at the very least, there should be a entry level boxed RH product that is stable and can be seen from a corporate IT standpoint as a branded RedHat product to deploy on their dns, ftp, http and other basic servers that may not be seen as warranting a Enterprise price tag. Otherwise, RH is going to lose big market share to Suse on this front. IMHO and all that. Hi Johnathan, It looks like not enough people voted with dollars and cents to ensure that RedHat could provide the services you're looking for. I see no reason to believe that Fedora will be less stable than RHL 10. If you need support you'll have to pay for it, the price is set by the market place. If someone else can provide what you need cheaper, by all means they're the correct choice for you.It would be great if RedHat could afford to be all things to all people, but we know that isn't realistic. The net result of this resent change may very well lead to more improvments to Linux, better support for the High end, and better community involvment in the low and mid markets. I just don't see any reason to worry yet. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 09:09:06 -0500 Mike Vanecek [EMAIL PROTECTED] wrote: Before I go on. I run RH9 at home. I used RH at the last place I worked. I do not like the behavior and configuration methods in Suse. SuSeConfig is an evil mess that can wax custom options on the next Online Update blah..blah.. blah. [snip] What options are left to the SOHO server user if not Fedora or SUSE? I've never used anything except RH, but would like to start thinking about a fall back plan in case Fedora is too bleeding edge for my needs. Both of these options are reasonably priced and give you more value, stability, and support for the money than you're likely to find elsewhere: http://www.redhat.com/software/rhel/es/ server for $349 http://www.redhat.com/software/rhel/ws/ workstation for $179 Good luck, Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 09:24:27 -0500 Ed Wilts [EMAIL PROTECTED] wrote: I've heard rumors that some announcements might be forthcoming this week. Let's be patient for a bit... Hey Ed, Care to share your source or to speculate what the announcement might entail? Redhat has already said that Fedora will _not_ be supported so all that i can imagine is some kind of price break on their enterprise offerings or as yet unmentioned mid-level product. Any Idea? Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 09:43:59 -0500 Dave Ihnat [EMAIL PROTECTED] wrote: On Tue, Sep 23, 2003 at 10:25:00AM -0400, Sean Estabrooks wrote: Both of these options are reasonably priced and give you more value, stability, and support for the money than you're likely to find elsewhere: http://www.redhat.com/software/rhel/es/ server for $349 http://www.redhat.com/software/rhel/ws/ workstation for $179 This is a joke, right? You ARE joking? Not at all. Do you think you DESERVE support for free, do you have any place to get the support they offer CHEAPER? Can you provide what they provide for a better price ? Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 09:05:28 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: Sean, in the corporate world you may be right, although even then, it strikes me that to take market share away from Windows, they should be cheaper. Given a two-year cycle of upgrades (not that common), it's $360 total which means it costs at least as much as Windows XP Professional, if not more. Same goes for servers, where we upgrade generally every three years: total cost of ES is $1,050 which was the cost of Windows licenses I was trying to avoid by moving servers to Red Hat. Well, you get a better product for that money but you're right it's not exactly cheap. I think it's _fair_ but i agree it might be out of range for some situations. However, that is not my primary point. What REALLY interests me is this: I, and thousands upon thousands of people, pay Red Hat $60/system/year for RHN service on RHL systems (either bought or downloaded). Typically we pay this for N machines, where probably 1 N 10. Together we are all a market which, in large part, simply cannot or will not afford WS, ES, AS, or even Mini-X. IMHO, Red Hat NEEDS to make sure that we are not forgotten or left out in the cold. You know that RedHat is bound by economic realities. I'm sure if there is a way to provide the services you want and turn a profit they will find a way to provide them. What exactly is it that Fedora does _not_ provide that you are willing to pay RedHat $60.00/server/year to get? Where, I ask, is RHEL-SS (SOHO Server)? Where is that $99/year price for SS which could serve as the first step for adoption of RH in corporate shops, or the only option for adoption of RH in very small shops? I WANT MY SOHO SERVER!!! Ed mentioned that perhaps such an offering is on the horizon. I'm struggling to imagine how they will differentiate it from Fedora and Enterprise. Editor's Note: I have a lot of faith in Red Hat, and I will give them every possible benefit of the doubt. I also recognize that the just launched Fedora Core YESTERDAY, and that we have worked ourselves into a lather in only 24 hours. So I will wait patiently for them to issue lots of other announcements which I'm sure will make things clearer. But today, at this point, so far, I do not see a product offering from RH which I can be _sure_ meets my needs and price point. THIS is my concern. Agreed. Regards, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 11:25:21 -0400 Buck [EMAIL PROTECTED] wrote: My thoughts too. I can just picture the following happening all over the country: I am a small business manager. I have 5 or more employees using XP at home. I want a file server for my network. Since I have a number of users that are familiar with Windows and the cost is less than Red Hat, Which server should I choose? If, on the other hand, I had an option of a Red Hat server under $200 with the automatic updates, and installation and configuration help, Red Hat would be nice. Just a little learning curve and I just need to add and delete users as necessary. Does this sound logical from that point of view? Sounds great unless it costs RedHat $201 to make it happen. If you can support yourself you can do it for $0. doesn't that sound better? You get the auto updating for that price too. Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 10:08:24 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: Rodolfo, I share your sentiment about RedHat. How's this for a thought? Make SS available for $99/year. Include everything a small business server needs (which I think makes it equivalent to ES), offer automatic updates via RHN, installed package tracking, plus a few (three, five?) incidents of WEB-ONLY support per year. No phone support at all. Unfortunately you've just described a product that offers MORE than Basic Edition ES. So how will they ever sell ES? I think this would be a good product for RH, and I _know_ I could then put at least $500/year into their pocket, if not more. How many thousands of people like me are out there? Care to help me make a little web page to count votes? Perhaps if RedHat just produced a Redhat-supported Fedora Release based on every second version of Fedora Core and only provided security updates this would satisfy you ? Maybe they could do something like that cheaply enough to provide the security to people who need it. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 14:31:20 -0400 Buck [EMAIL PROTECTED] wrote: Support What support? The products you listed are only the product and up2date for 1 year! The cost of these two products is actually $799 and $299 respectively. Look again my friend. Actually you're wrong Buck. Go look. We're not talking about telephone suport. Nobody was getting telephone support for $60 dollars a year. Buck. Please, tell me what it is that you're willing to pay for that is _not_ provided by Fedora? More importantly tell me what it is in the Enterprise options you're not willing to pay for? They have to decide their business plan for themselves and their shareholders. I think they're making good decisions. RedHat is NOT going to drop their Enterprise prices down. There is a market willing to pay those prices and recognise the VALUE. i'd be impressed if someone here can articulate what they're willing to pay for above Fedora and not willing to pay for (ie. can be removed) from the Enterprise offerings. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 15:30:54 -0400 Buck [EMAIL PROTECTED] wrote: My sentiments exactly. But add this. If I can support the product, I get a customer. Along with that, I would purchase a boxed edition for the customer to put on his shelf, and at least a minimal support package so I can call for help should I need it. I think it would be great if the Server package were available with a basic installation and configuration guide and 2 months of up2date included for $100.00; Additional support packages for $100 per hour/incident -- prepaid gets priority support; and for $50-75, a years Up2date extension. For less than the price of a Microsoft server, I could sell them installation and support their server for a year. Better bargain than MS. Maybe that's the ticket. A boxed set of Enterprise with 1 month up2date. Could be a way to amortize the cost of their boxed set production across more offerings? Perhaps RedHat could start a VAR program. Might be an option to reduce the costs, and service the low end. RedHat has to be careful that they don't drain their resources or hurt their sales of Enterprise. There's obviously a niche below Enterprise, the question is, can RedHat make a profit servicing it. Will be interesting to see if they make some offering and what the mix is if they do. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 13:54:32 -0700 Rhugga [EMAIL PROTECTED] wrote: Yea, I have to agree with most on this thread, there is no future in selling Red Hat/Linux support. It has already been tried anyway on the west coast with these companies that started competing with Sun/EMC/HP/Compaq support. (I was actually just contacted by a new one yesterday) I think VA-Linux had a great business model, just 3 years to soon. Hey Chuck, I take it you've _never_ purchased _anything_ from RedHat? After all you don't need the Media, you can just download it. And you don't need the support, you get that for free too. Why would you ever send any money to RedHat? Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 16:49:27 -0400 Buck [EMAIL PROTECTED] wrote: My answer you quoted below are in response to a statement that WS is only $179 with support and ES is $349 with support. I corrected him by saying that the prices with support are $299 for WS and $799 for ES. Nobody is talking about Telephone support which is what the prices you quoted include. The legitimate concern that i see people voicing is that Fedora does not have an appropriate Support/Updates cycle for business use. While Enterprise doesn't have this problem it's too expensive for some uses. Enterprise has a 5+ year cycle and Fedora has a 3+ month cycle. Maybe a product with a life cycle somewhere in the middle could be offered at a reasonable price. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Tue, 23 Sep 2003 15:17:30 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: 1. I would like updates and patches to be issued for longer than 12 months, maybe 18 or 24 so I don't have to reinstall as often. Sure, this is a real gap in the product range. 2. I do not need any support at all... none whatsoever. Just the updates, ma'am, delivered automatically by some software certainly have more than one that can do the job). In my vernacular, that _is_ support, but i accept your usage. IIRC, this is more than Fedora but less than RHEL. But for this, I would be willing to pay $0 X $180 per year. The challenge for RedHat is to provide a product that doesn't step on the toes of its enterprise offering. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Wed, 24 Sep 2003 10:08:58 +1000 Ian Mortimer [EMAIL PROTECTED] wrote: A large and potentially larger part of the Linux market is in the education sector. Many schools/colleges/universities don't have budgets to pay for expensive software licenses (one reason many are moving from Windows). Some software vendors offer unsupported non-commercial licenses at discounted rates or even free in recognition of the promotional advantages of having their products in the education sector. It seems to me that unless RH do something similar admins in this sector (even those with an RHCE!) are likely to start looking at other distributions (debian is already popular - partly because of knoppix) or*bsd. Tom Callaway of RedHat already said that they'd be open to helping out: http://www.redhat.com/archives/fedora-list/2003-September/msg00133.html Regards, Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: PAM News
On Wed, 24 Sep 2003 09:26:55 +0800 Edward Dekkers [EMAIL PROTECTED] wrote: Michael Schwendt wrote: Check the real log files in /var/log, not the stripped-down information presented to you by logwatch. Also check whether it's a daily cron job in /etc/cron.daily OK, the real log files tell me the job was run at 4:08.31 and 'su news' was executed then, and logged out one second later. There IS a job in cron.daily called inn-cron-expire. This seems to be causing it. Not being a bash expert - what does this thing do? Anybody know? It seems to check chkconfig to see if innd is active, then performs the su. However, innd is NOT active on my PC (I checked with chkconfig --list, it's off on all runlevels), so why does it still perform the su? Can someone (who understands bash)check theirs and make me understand? Ed, The script essentially just checks if inn is installed. It doesn't test to see if it is disabled on all run levels.You can uninstall the inn package if you're not using it. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: XFree86
On Tue, 23 Sep 2003 17:38:08 -0800 Noah [EMAIL PROTECTED] wrote: can somebody provide clues for a method for troubleshooting this issue: the /var/log/XFree86.0.log is attached. I hope you can see that too. # startx XFree86 Version 4.2.1 (Red Hat Linux release: 4.2.1-21) / X Window System (protocol Version 11, revision 0, vendor release 6600) Release Date: 18 October 2002 Build Operating System: Linux 2.4.21-1.1931.2.274.entsmp i686 [ELF] Build Host: daffy.perf.redhat.com Module Loader present OS Kernel: Linux version 2.4.20-20.8 ([EMAIL PROTECTED]) (gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7)) #1 Mon Aug 18 14:42:17 EDT 2003 Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: /var/log/XFree86.0.log, Time: Tue Sep 23 18:23:42 2003 (==) Using config file: /etc/X11/XF86Config-4 Fatal server error: AddScreen/ScreenInit failed for driver 0 - X server log file: /var/log/XFree86.0.log - Kernel log file: /var/log/messages XIO: fatal IO error 104 (Connection reset by peer) on X server :0.0 after 0 requests (0 known processed) with 0 events remaining. Hey Noah, This is a bit of a generic error when the server can't start for some reason. Something is misconfigured in /etc/X11/XF86Config You'd have to look in the log files mentioned to determine exactly what the problem is. Instead you could just try running this command: X -configure This will try to create a proper config file but it won't replace your existing config. To test the generated config you run: XFree86 -xf86config /root/XF86Config.new If the server starts without a problem you can replace the /etc/X11/... file with the generated config and you should be away to the races. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Fedora
Just discovered this announcement from earlier today and it looks to be worth sharing. http://www.fedora.us/ http://fedora.redhat.com/ http://fedora.redhat.com/about/rhel.html Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On 22 Sep 2003 11:29:48 -0700 Rick Warner [EMAIL PROTECTED] wrote: On Mon, 2003-09-22 at 11:23, Benjamin J. Weiss wrote: So, now I'm confused. Does this mean that I won't be downloading RH 10, but instead will be downloading Fedora 10 or something? My reading is the Fedora is to RH Enterprise Server as Rawhide is to RedHat Linux. It is a developer supported testbed for things that may end up in future releases of ES. Seems to be distinct from RedHat Linux that has traditionally been available for download. Hey Rick, My reading of it is that the Fedora project is being merged with the future of the non-enterprise version of Linux. There won't be any need for RedHat to release another non-enterprise version of Linux. Fedora is it. You're right though, i'm sure RedHat justifies their involment with it as a way to feed their Enterprise offering. Not sure what will become of Rawhide, perhaps they'll reuse that name for -test- versions of Fedora. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 13:06:54 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: No, do not mix in Rawhide. Rawhide is the kitchen, where they put stuff now being tested or worked on. It is always going to be, well, raw. Rawhide is not a release of any kind. Fedora Core replaces Red Hat Linux. It is now a more open programming model than RHL was, and so the community will have more involvement and there will be more packages. Red Hat, Inc. will put a lot of effort into helping Fedora Core grow and improve, the same kind of effort that they put into RHL, but Fedora Core will not be a branded RH product. RHEL is Red Hat's primary commercial product which they hope will drive they revenues, and with which they hope to compete head-on against Microsoft in the corporate space. Well said. All i'd add is that it _will_ be a branded RedHat project, there will be no product as such. From the Fedora site: Fedora is now a trademark of Red Hat, Inc. Red Hat will defend this trademark in order to protect the integrity of The Fedora Project. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 14:13:12 -0500 Benjamin J. Weiss [EMAIL PROTECTED] wrote: Okay, now for the big question: RHCE? As I understand it, the current RHCE exam is based on RH 8.0. If RedHat is now dropping the free version of RedHat in favor of the Fedora Project, is the RHCE going to be splintered? Or are they going to change the test to be based on the Enterprise Linux? Now I know I must be getting old, I'm getting tired of all the change that abounds...*sigh* Lol... hey Ben, Pretty safe to say that the RHCE will be based on the enterprise offerings. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 14:19:49 -0500 Benjamin J. Weiss [EMAIL PROTECTED] wrote: I hope that RH isn't shooting themselves in the foot with this. I think I understand the rationale, but it was the close branding between Red Hat Linux and Red Hat Enterprise Linux that convinced my boss to go ahead and purchase RHEL Advanced Server. I'd already deployed a couple of RH 9 boxes and showed him what RH was capable of, and when he found out that there was a commercial version that had enterprise support, he decided to purchase. He viewed RHL as the entry level and RHAS as the professional/commercial level. I don't think that he's anywhere near alone in that view. I'm afraid that by making this move, RH may do two things: 1) Fedora won't be as popular with the newbies, since it won't have the famous Red Hat name, and the distribution will lose newcomers to Suse, Mandrake, and Debian. 2) RHEL will lose market share because they won't have the hordes of people trying/using RHL at home or for small project who then look to upgrade to RHEL. I'm worried, I am. As far as i can see the only real difference from what we have today is: a) You can't buy a boxed set of Fedora b) You can't buy support for it either. c) The average developer can be more involved. Other than that, everything else will be the same. P.S. What'll happen to RHN for the free linux?? You'll have to purchase an Enterprise offering to get support from RedHat. But there will surely(?) be auto updating feature provided for Fedora. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 12:34:33 -0700 Samuel Flory [EMAIL PROTECTED] wrote: You'll have to purchase an Enterprise offering to get support from RedHat. But there will surely(?) be auto updating feature provided for Fedora. Fedora currently supports yum, and apt-get. I've been using fedora, and before that freshrpms with apt-get, and yum for 6 months. I've been very happy with them. Hi Samuel, I haven't been using them that long but have liked what i've seen, especially Synaptic. I am hopeful that this service will continue. I guess that's the part that Fedora really brought to the table in this merger. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 15:45:02 -0400 (EDT) Gerry Doris [EMAIL PROTECTED] wrote: At 12:23 9/22/2003, you wrote: Does this mean that I won't be downloading RH 10, but instead will be downloading Fedora 10 or something? Fedora Core 1 (Cambridge), apparently, which will contain everything you expected to see in Red Hat Linux 10 and more due to the contributions of the Fedora Project. and less since Redhat will insist on several apps being removed due to potential licensing issues. Gerry, Nothing will be removed from what was expected in RHL 10. Nobody expected to see MP3 support in it for example. Some things that have been part of the Fedora Project will have to be pushed out to new homes because of Licensing or Patent issues. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 14:52:52 -0500 Benjamin J. Weiss [EMAIL PROTECTED] wrote: I hate to say it, but I'm a bit lazy. I *love* being able to do an up2date -u and have all of my out-of-date RPMs updated. Will Synaptic or apt-get do this? Yes, and Synaptic is a very nice GUI that'll let you use your mouse to be more selective as well. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 15:51:12 -0400 Fournier, TJ [EMAIL PROTECTED] wrote: Which apps may those be? TJ Fournier I don't know the comprehensive list, but things like MP3 support, mplayer, Valgrind, etc, etc... Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 14:59:48 -0500 Benjamin J. Weiss [EMAIL PROTECTED] wrote: 3. Will the Fedora Project Releases be compatible to existing Red Hat? In other words, can I upgrade what I have to Fedora or will I have to start with Fedora from scratch? You can upgrade to the current Red Hat beta via apt-get or yum using fedora now. All I did was the following: Hey, another thoughtare we going to have to change the list name to Redhat and fedora list? ;p They've already set up new lists: http://fedora.redhat.com/participate/communicate/ * fedora-list - For users of Fedora Core releases * fedora-test-list - For testers of Fedora Core test releases * fedora-devel-list - For developers, developers, developers * fedora-docs-list - For participants of the docs project Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Fedora
On Mon, 22 Sep 2003 15:25:15 -0600 Rodolfo J. Paiz [EMAIL PROTECTED] wrote: At 13:30 9/22/2003, you wrote: As far as i can see the only real difference from what we have today is: a) You can't buy a boxed set of Fedora b) You can't buy support for it either. c) The average developer can be more involved. Other than that, everything else will be the same. No, Benjamin has a point: another change (d) is that the free version will not be branded Red Hat, as I also mentioned earlier. This _could_ believably slow down adoption due to brand recognition, both in the case of newbies choosing the well-known Red Hat for their Linux, but also of people like Benjamin's boss (and mine, by the way) who choose or would choose RHEL after seeing RHL at work, but who will not see Fedora in the same light. Perhaps, but i think most people are sophisticated enough to add RedHat in front of Fedora for themselves. Fedora is now a trademark of Red Hat, Inc. Red Hat will defend this trademark in order to protect the integrity of The Fedora Project. Also, there may very well be CD's made avaialble of the Enterprise versions so you can install those as a first step. They wouldn't be distributed by RedHat, but would still be free (as in beer speech) if you didn't require support. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: routing challenge
On Sun, 21 Sep 2003 11:28:48 -0500 MKlinke [EMAIL PROTECTED] wrote: On Saturday 20 September 2003 08:38, Asbjorn Hoiland Aarrestad wrote: AFAIK bridging isn't an option on wireless, at least not for this kind of relay network, if it were then it would be possible to treat all the nodes as a single segment with no need for layer 3 routing. But then again the subject line did suggest a routing question not a wireless question ;o) my experience with wireless networks aren't that big either, and that's why I put this question here. The subject line was chosen from my networking experience. What kind of bridging are you mentioning here? Is it possible to make this network work without having to get into complex routing tables? - asbjørn Take a look at: http://www.docs.uu.se/docs/research/projects/selnet/lunar/ as this sounds like it may work for your project if the capability of the wireless cards you're using are OK. Hey Mike, Nice find, it looks pretty good. It pushes the routing discovery to a custom layer below IP in a pretty clever way. One word of warning though, they don't recommend networks with more than 3 hops so it may not meet the goals of this particular project. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: routing challenge
On 19 Sep 2003 17:28:23 -0500 Bret Hughes [EMAIL PROTECTED] wrote: I have zero experience with wireless so excuse my boneheaded question(s). I think this is going to make me take an orthogonal view from how I normally picture a network. Perhaps it is my needing to read up on routing daemons that would fill in my blank spots. Hey Bret, I have only slightly more experience with wireless networks and because of the complexity was only trying to nudge the OP in the right direction rather than provide a solution. The solution i did hint at wasn't meant specifically for wireless, same thing can be setup on a local lan. For instance in a classroom situation each node is configured with multiple logical networks on the same physical lan. Even though all the packets are flying across the same physical media, only the proper network sees each packet because of the logical segmentation. This was the genesis of the notion that i described. AFAIK bridging isn't an option on wireless, at least not for this kind of relay network, if it were then it would be possible to treat all the nodes as a single segment with no need for layer 3 routing. But then again the subject line did suggest a routing question not a wireless question ;o) My usual thinking ( forget wifi for a minute) base server node1-node2 ... node12---remote server in a wired scenario each node would have 2 nics and each segment between the boxes would be separate networks with exactly 2 hosts. 4 ipaddresses would be required to correctly subnet each network. Now, Sean said make each node a network and since there is only one interface that can talk to multiple networks, this is where I get bumfuzzled since as I said, I usually thing of the link between machines as the network. From each nodes perspective it is on a wireless lan consisting of just its immediate neighbors. The problem becomes how to route packets between these smaller overlapping physical networks. Ideally a solution would route packets as far as possible in each hop and be robust enough to adapt if a node goes down. example: All of the nodes are on the same wireless subnet. However each node is also configured internally with it's own unique network. For example node4 also responds to IP's 10.10.4.1 and is the only node on the 10.10.4.0/24 network. (I suspect this is the point that wasn't clear in my original post) wireless lan: 192.168.0.[1-5] one for each node unique network IP:10.10.[1-5].1 one for each node base1---+-+ node2---+--+ node3---+-+ node4--+ remote5 (diagram sux unless viewed with a fixed width font) base1 routing table: dest gateway 10.10.2.1 192.168.0.2 10.10.3.1 192.168.0.3 10.10.4.1 192.168.0.3 10.10.5.1 192.168.0.3 node3 routing table: dest gateway 10.10.1.1 192.168.0.1 10.10.2.1 192.168.0.2 10.10.4.1 192.168.0.4 10.10.5.1 192.168.0.5 When base1 wants to send a packet to node2 it _could_ use 192.168.0.2 directly because that node is in wireless range. But what if it wants to reach remote5? 192.168.0.5 is NOT in wireless range. So instead of sending to 192.168.0.5 we send to IP address 10.10.5.1. At this point the routing tables take over and forwards the packet to node3 (192.168.0.3). Node3 has a direct route entry so forwards the packet again to 192.168.0.5 (from here 192.168.0.5 is in range!) When the packet arrives at 192.168.0.5 the node recognizes that requests for 10.10.5.1 are for itself and processes it locally. So the answer is to _always_ use the fake network/node number rather than the real wireless lan IP, routing will take care of the rest. While you could do this all with static routes it's not as flexible. So the only remaining setup is to use a routing daemon to configure the routing tables so that they are adaptable to outages and the like. On a good wireless day with little interference perhaps the routing tables will show that node4 can be reached directly from base1 without a need to route through node3. Other days packets destined for node4 from base1 will have to route through both node2 and node3. I think of this as logical links (assume that each machine can see the next two) / ---\ base server node1-node2 ... node12---remote server \--/ Since the network interface on any given node can see multiple networks is this tantamount to running multiple subnets on the same wire? Is there the concept of a default gateway in this scenario? I can't see how since in all neworking I have done the next hop or default gateway to another network is on the same subnet as the host. No need for a default gateway, although it doesn't hurt to have one. The
Re: routing challenge
On Sat, 20 Sep 2003 15:38:55 +0200 Asbjorn Hoiland Aarrestad [EMAIL PROTECTED] wrote: AFAIK bridging isn't an option on wireless, at least not for this kind of relay network, if it were then it would be possible to treat all the nodes as a single segment with no need for layer 3 routing. But then again the subject line did suggest a routing question not a wireless question ;o) my experience with wireless networks aren't that big either, and that's why I put this question here. The subject line was chosen from my networking experience. What kind of bridging are you mentioning here? Is it possible to make this network work without having to get into complex routing tables? Hi Asbjørn, Just meant generic layer-2 bridging so that you didn't have to worry about network layer routing. Was really thinking out loud and hadn't thought it through completely. Anyway, I seem to recall that there is a limitation on most(all?) wireless cards which precludes their use in a bridging setup. The more i think about this option the more complex it seems. As for the routing setup i outlined you're right that it is a tad complex, but it in the end it seems pretty straight forward. To get started you could make all the routing tables static and only add the routing protocols/daemons later to improve reliability. You could test everything out with a few desktops and/or laptops to make sure you had it sorted out before you tried to implement it over any real distances. Anyway, if you pursue this or a different option it would be great to hear from you on your progress and success. Good luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: pinging a name vs. number
On Fri, 19 Sep 2003 10:59:57 -0400 Timothy Stone [EMAIL PROTECTED] wrote: List and network gurus... I recently had a coworker tell me that we should *always* ping by *name*, *not* by number. The reason as explained is because the resolution takes place at a higher level on the OSI stack. And pinging the number does nothing but tell you that the 'connection' to the machine is okay. Huh?! Isn't pinging by name just the same as pinging by number, only adding a step for the name resolution? You're correct. Pinging by name will also test name resolution but the connectivity test is exactly the same. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: vim-6.1 syntax highlighting
On Fri, 19 Sep 2003 11:49:53 -0500 christopher j bottaro [EMAIL PROTECTED] wrote: the debian systems at my school have vim-6.1 but don't do syntax highlighting. how do i get syntax highlighting like on my redhat-9 system at home? Hi Christopher, If the syntax highlighting files are installed but just not being used, it might be as simple as typing: :syntax enable If they aren't installed or VIM has been compiled without syntax highlighting support you'll have to do some reinstallation work. The syntax files which describe what to highlight for each file type are in $VIMRUNTIME/syntax which on redhat 9 is: /usr/share/vim/$VERSION/syntax The online help for vim is really quite good. When using vim type: :help syntax and you'll get oodles of info. Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: script to monitor diskspace
On Fri, 19 Sep 2003 08:41:11 -0400 Marvin Blackburn [EMAIL PROTECTED] wrote: Does anyone have a shell script or procedure to monitor disk space. i.e. notify someone when a filesystem gets too big? Lots of ways to do this of course. One simplistic way is to add the following line to one users crontab file*: 59 * * * * df -h | awk '$5 80 NR 1' The user will be emailed a report only if any filesystem has greater than 80% usage. In the above example, the usage check is made every hour.If this user has a .forward file in their home directory the email can be sent wherever they like. One downside to this method is that the report will be sent _every_ hour until the usage is reduced. HTH, Sean * crontab -e is an easy way to update a users crontab. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: routing challenge
On Fri, 19 Sep 2003 10:14:40 +0200 Asbjorn Hoiland Aarrestad [EMAIL PROTECTED] wrote: Got a challenging routing problem, and perhaps some of you can help me. The network is as follows. We have 12 standalone nodes with wireless lan cards. The nodes are placed on a line, and the average distance between each node is 400 meters. With good wireless cards, that means that we can reach two or three node neighbours on each side of a node. (the nodes will be placed in the middle of nowhere, security is not the issue since the biggest security risk will be the ice and polar bears. This also means that putting up tp-wires is not an option) on each side of the line of nodes, there will be a server. There will only be people at the first server (beside the first node) The nodes will be collecting data, and we want to transfer the data to the server beside the first node. But we also want to be able to telnet into each of the nodes to be able to do maintainance (so we don't have to use a snow-mobile and get very cold each time there is a problem on one of the nodes) Any ideas on how to make a good enough network, enabeling us to communicate between the nods and the servers? Hi Asbjorn, Sounds like a fun project. The first thought that springs to mind is that you treat each node as a separate network. Essentially a network of 1 with each node having a fixed IP address on its own subnet. Then run a routing daemon on each node that discovers its neighbours and shares that route information with each of them. This should give you some resiliency in the face of a single node dropping out. Obviously this is rather too complex to describe in detail here but you should take a look at routing daemons such as Gated or Zebra. Zebra is easier to setup and should do everything you need. Actually Zebra has recently been forked as Quagga so the place to start reading might be: http://www.quagga.net/ Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Abiword 2.0.aaarrrrrghghghhghgh
On Fri, 19 Sep 2003 14:26:15 -0400 mark [EMAIL PROTECTED] wrote: On Friday 19 September 2003 02:06 pm, From: Michael Lee Yohe [EMAIL PROTECTED] Why are you using a SuSE package? Instead, simply download Abiword-2.0 from Rawhide. I *do* wish folks would read what others wrote. Mark, You're obviously frustrated but I think you should read what Michael wrote. If you noticed, I *also* tried the abiword-gtk* package, and the abiword-gnome* package, which I didn't mention. That's three seperate packages, and all of them had dependency problems where there should not have been. None of those have anything to do with rawhide, which is what Michael said. Why haven't you tried the rawhide packages? Took about 60 seconds to install abiword here. rpm -ivh abiword-2.0.0-2.i386.rpm fribidi-0.10.4-3.i386.rpm \ libwpd-0.6.2-1.i386.rpm Btw, *if* I read the info at rpmfind correctly, the frbidi0 package is supposedly only for *development, and shouldn't be needed, only the frbidi package should be necessary. And why can't it find libcrypto and libssl? Perhaps because you're trying to use rpm packages from another distribution rather than those released from RedHat ? mark, looking for answers to what he asked, not what words someone put into his post You really need to take your own advice. Even if Michael had made a mistake, which he didn't, a little grattitude for his attempt at helping you would be in order. Regards, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: RedHat 9 rc.sysinit problem
On Fri, 19 Sep 2003 12:57:08 -0700 Jim Dickenson [EMAIL PROTECTED] wrote: I did a clean install of RedHat 9 on a system. This system has only a SCSI drive. I was able to reboot the system with no problem. I then applied some updates and did some other stuff and when I tried to reboot the system the boot process stopped when trying to run fsck on the root, and only partition. The problem is that the partition is mounted read-write by the time rc.sysinit is called so fsck gets an error that causes rc.sysinit to call sulogin. I did a goggle search and noticed one other person had a similar problem but no one responded to his request for an explanation. Here is the part of rc.sysinit that is causing the problem: if [ -z $fastboot -a X$ROOTFSTYPE != Xnfs ]; then STRING=$Checking root filesystem echo $STRING initlog -c fsck -T -a $fsckoptions / rc=$? Can someone tell me what might have caused this problem and/or what a solution to the problem, other than not running the fsck, might be to fix it? HI Jim, The root filesystem should be mounted read-only when rc.sysinit is first called. If it isn't, this suggests there is a problem with your /boot/initrd-*.img file.This image contains a script that is responsible for mounting the root filesystem as read-only. Try rebuilding it with themkinitrd command. Also, make sure that it is properly referenced in your /boot/grub/menu.lst file. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: rescurive ls to find symbolic links
On Fri, 19 Sep 2003 15:25:42 -0500 Rigler, Steve [EMAIL PROTECTED] wrote: To find symbolic links recursively you could do: find . -type l -print I don't know of a way to find hardlinks. Hard links are indistinguishable from the original file. Whereas the softlink search you showed won't return the original file the following search will return the hardlinks _and_ the original file: find -type f -links +1 Unfortunately this won't group hardlinks associated with each file together, for that you need something like: find -xdev -type f -links +1 -printf '%i %p\n' | sort -n Which prepends the inode (unique file number) in front of each hardlink so that they can be properly grouped by the sort command. The -xdev option is needed so that only one filesystem is scanned because inode numbers are not necessarily unique across filesystems. To combine the hard soft link search into one: find -type l -o -type f -links +1 Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Windows Key
On Fri, 19 Sep 2003 11:53:37 -0700 Robert Denton [EMAIL PROTECTED] wrote: Hello all, what is the easiest and/or best way to make the Windows Key useful in RH9? I would like to use it with 't' to bring up a gnome-terminal. Any suggestions? Robert, The following commands will configure Metacity to recognize the Windows key (mod4) + T and tell it to open gnome-terminal: gconftool-2 -s /apps/metacity/keybinding_commands/command_2 \ -t string gnome-terminal gconftool-2 -s /apps/metacity/global_keybindings/run_command_2 \ -t string mod4t The first command says what to run, the second assigns the keystrokes. You just have to run these once as the user you wish to configure. You'll notice command_2 in each of the instructions above. You can configure 12 different commands for Metacity in this way (_1 through _12) but _1 is preconfigured on RH9. Alternatively, you can usegconf-editor instead of the command line version above. It is very similar to the MS Windows registry editor. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: creating multi volume tar files .tar1, .tar2 etc
On Wed, 17 Sep 2003 02:02:25 -0700 Ian L [EMAIL PROTECTED] wrote: Hey Ian, well it seemed like a bad idea to commit ourselves to using a backup drive for which buying new tapes would probably not be possible, plus if the drive bit the dust, we wouldnt have access to any of our backups. It seemed like a really nice drive. I was hoping they'd recover, but doesnt look like it. This seems like a smart business decision, given the critical nature of backups. how do i rejoin the tar files? make one big file out of them :cat tarfiles* big.tar or to extract them directly : cat tarfiles* | tar -xvf - I might approach this differently. Right now i have one machine which is acting a backup server. I have a few other machines which mount to a directory on the backup server. To avoid this 2gb limit (upgrading nfs on multiple machines all running different version of redhat doesnt seem feasible :) i think i might mount all the machine directories directly on the backup server, and run tar from there. That'll avoid the file size limitation since i wont be writing to an NFS mount that way. If you're _reading_ across NFS you'll still bump into the limit. RedHat 9 has v3 and IIRC so do all RedHat releases going back to 7.1. You might want to look at rsync to reduce the amount of network traffic needed to do a backup, or even consider a network backup solution like Amanda. Good Luck, Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Samba Question: Want to enable specific users from windows domain to access samba share.
On Wed, 17 Sep 2003 13:23:40 +0300 [EMAIL PROTECTED] wrote: Hi All, I want to enable two users from our windows domain to access particular share from my samba server. I am new to linux ... please help me out. I'd setup public share and configured samba long time before. It works fine. Now I want to create new share to give access to particular users. Here is how I did that. * I created a folder, lets say folder in my Linux box. * Change its permission to 0765. * The user names of the windows network are 14200 and 14210. * I tried # smbadduser 14200:My_name * It gives me error 14200 is not in PASSWD database SKIPPING... * Then I tried to create a user with # adduser 14200 * It gives the error not a valid name. (It seems to me that Linux does not support only numeric characters in user name). Hi Yasir, You are correct that user names should not start with a number. Fortunately Samba allows you to map SMB names to local user names. You were on the right track with smbadduser except that the local name has to come first. smbadduser My_name:14200 also make sure this line is uncommented in your smb.conf file: username map = /etc/samba/smbusers This tells Samba to use the mappings found in the smbusers file which looks like this: Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: upgraded ssh from 3.1 to 3.7.1 - now getting connection refused
On Wed, 17 Sep 2003 11:45:23 +0100 Martin Moss [EMAIL PROTECTED] wrote: I just upgraded my ssh using the 3.7.1 source tarball (I couldn't find an rpm for it). Now when I try to login I get a connection refused. As I am unable to get a connection to the machine, I am not able to provide much debugging information. Can anybody give me any pointers as to what to look for to sort out the problem? I do have a go between who has access to the console. the machine in question uses an iptables firewall, but prior to the upgrade the system worked fine letting connections to port22 through. Martin, I'd suggest reinstalling the 3.1 version from rpm, at least until you can sort problem out in a computer that is closer to you. Trying to debug through a remote third party with help from the list is going to be difficult. What problem were you trying to solve by upgrading to 3.7 in the first place? Perhaps, upgrading to the current rawhide rpm package (3.6) would be easier and still provide what you need. Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: upgraded ssh from 3.1 to 3.7.1 - now getting connection refused
On Wed, 17 Sep 2003 12:23:15 +0100 Martin Moss [EMAIL PROTECTED] wrote: There is an exploit for all versions lower than 3.7.1. Allowing root access. It seems to have hit the public domain yesterday. http://www.linuxsecurity.com/advisories/redhat_advisory-3628.html (this one seems to say 3.1 is ok, but When I first started hunting another site I came across (which I now can't find gr) was suggesting that 3.7.1 should be the version to install, which I did. Martin, Thanks for mentioning the security alert. Yeah, like you noticed the page you referenced gives you RedHat rpms that are said to close the secuirty problem. Think you found your solution :o) Here is the same information from the RedHat site itself: http://www.redhat.com/archives/redhat-watch-list/2003-September/msg3.html Good Luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Serial ports - how do you verify?
On Wed, 17 Sep 2003 08:04:10 -0400 Keith Birchfield [EMAIL PROTECTED] wrote: Can anyone suggest a method to verify that my serial port is properly installed and functioning well on Red Hat 9? I'm getting inconsistent reads on a software dongle. You can see some information about your serial line settings with commands like (change device name to suit): setserial /dev/ttyS1 stty -a /dev/ttyS1 This won't tell you much unless you know what your are looking for. What settings does the software recommend for access to the dongle ? Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: RedHat 9 proc/kcore broken
On Tue, 16 Sep 2003 09:24:41 +0100 Anthony Hay [EMAIL PROTECTED] wrote: Anyone know why I can't see the PC BIOS at F hex in proc/kcore? Hi Anthony, /proc/kcore is not a raw memory image, it's actually in the ELF object format. Its header is 4K which displaces the bios, so you'll find the BIOS at file offset F1000. The following will extract the bios data from /proc/kcore: dd if=/proc/kcore of=bios_kcore bs=1024 skip=964 count=64 to extract it more naturally use /dev/mem instead: dd if=/dev/mem of=bios_mem bs=1024 skip=960 count=64 Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Watching an ssh version
On Tue, 16 Sep 2003 08:02:39 -0400 Reuben D. Budiardja [EMAIL PROTECTED] wrote: Hi, I remember this is being asked here long time ago, but I couldn't find it in the archive, probably due to inaccurate search terms. Anyway, I am wondering if there is a way to watch an SSH session. Suppose a friend of mine is accessing my computer remotely to do some configuration via SSH. Is there a way for me to literally watching him from local to see all that he is doing on the terminal, comand by comand? Suppose both of us have root access. This is not because I don't trust the person, but for learning purposes. Hi Reuben, This solution doesn't really have anything to do with ssh but should work for you. Log in as root and start the screen text window manager: # screen Once your friend is connected as root via ssh they type: # screen -x This will connect both screens together. Either of you can type and both will see the result of any commands executed. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: OT? Kernel Tuning
On 16 Sep 2003 15:57:11 -0400 Jason Dixon [EMAIL PROTECTED] wrote: Hey Jason, Since this thread was already appropriately labeled off-topic thought i'd pipe in. ;o) On Tue, 2003-09-16 at 15:30, Kelerion wrote: nice answer.. now thats what I like to see in mailing lists.. Thanks. informative.. helpful.. but leads the OP to figure things out him/herself.. there should be more of this on mailing lists... Unfortunately, there are too many a) folks looking for the quick answer. These are, more often than not, people raised on the Windows way of doing things... call support, get put on hold, get told which The List is a great forum for quick answers to problems that people are faced with. How many times have you yourself looked for an answer on the list? Clearly you know how to _fish_ so you're not looking for someone to say RTFM. If there is an online reference to more information that's a great thing to include in an answer, but why make a paragraph out of an answer that can be stated with a quick example? Often a person who is just learning about Linux does not have enough context yet to appreciate some of the subtleties of a solution. In these cases especially, a quick answer solves their immediate problem and should provide inspiration for them on how to proceed in the future. buttons to press. Or b) too many folks throwing easy answers out there without explaining how the solution was resolved. This is also a problem with many Linux HOWTO documents. I really enjoy sharing the knowledge that I've acquired over the years. However, this rarely comes in the form of you need to set SHMMAX to blah value, or compile dillywankers into APCI support. More often than not, it's go here, check this out, see where it leads you. This is more efficient for me, and, for the self-serving student, a much better opportunity for long term retainment. If the solution to a problem is to set SHMMAX to blah value then that is the proper answer to a question.If you have the time to explain the highlights of why it works that's even better. The correct answer should lead them to the correct source for more information. If it's not obvious where this information can be found it's great if you can cite a reference for them. For example if you tell someone that the answer to their problem is a specific iptable command, then clearly they will know they need to read more about iptables. No need to tell them to do that in each and every post. Not to mention, they're more likely to offer the same assistance to others later on. :) IMHO, more people are served by seeing the correct answer. They're already looking for the correct answer, just give it to them. Don't assume someone asking a question is stupid and needs to be told to go do more research. Obviously when you've asked questions on the list you were looking for an answer not someone to teach you how to fish. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: OT? Kernel Tuning
On Tue, 16 Sep 2003 17:14:00 -0700 Chris W. Parker [EMAIL PROTECTED] wrote: Hey Chris, Sean Estabrooks mailto:[EMAIL PROTECTED] on Tuesday, September 16, 2003 4:56 PM said: If there is an online reference to more information that's a great thing to include in an answer, but why make a paragraph out of an answer that can be stated with a quick example? Because it's more informative and helpful. Sure in certain cases. But often it's more helpful just to give the command. Often a person who is just learning about Linux does not have enough context yet to appreciate some of the subtleties of a solution. In these cases especially, a quick answer solves their immediate problem and should provide inspiration for them on how to proceed in the future. Should, but doesn't necessarily. That's really up to the poster. If the solution to a problem is to set SHMMAX to blah value then that is the proper answer to a question.If you have the time to explain the highlights of why it works that's even better. Yes, and that's what he did. Wasn't talking about this case, was talking about the general statements Jason was making. The correct answer should lead them to the correct source for more information. If it's not obvious where this information can be found it's great if you can cite a reference for them. For example if you tell someone that the answer to their problem is a specific iptable command, then clearly they will know they need to read more about iptables. Clearly? Yes, clearly. No need to tell them to do that in each and every post. For that to be the case we'd need to have some sort of collective mind, or hive mind (ala the Borg) where eveyone knows exactly what everyone else knows. You learn something, I learn it too. Unfortunately this is the real world and your statement can't possibly be true since people continuously ask the same questions. Not to mention, they're more likely to offer the same assistance to others later on. :) IMHO, more people are served by seeing the correct answer. This is true. Wow, we agree on something ;o) They're already looking for the correct answer, just give it to them. Don't assume someone asking a question is stupid and needs to be told to go do more research. Obviously when you've asked questions on the list you were looking for an answer not someone to teach you how to fish. Man I'm totally confused by your post. The original poster ASKED to be taught how to fish INSTEAD of being thrown a fish. Again, talking about his general statements, not this case. Am I completely missing the point of your post? Apparently ;-P So confus-ed!! Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: OT? Kernel Tuning
On Tue, 16 Sep 2003 17:39:52 -0700 Chris W. Parker [EMAIL PROTECTED] wrote: Chris But how do you decide what's best for the poster? How do you know that a simple answer is more helpful to them than a more lengthy one? Unless they tell you outright it'd be really difficult to determine. This isn't really a question about length of answer, more about style. My personal preference is to give as much explanation as is necessary to get started but not to try to write a book. And when an example can _replace_ a paragraph, then IMHO the example is the correct choice. This just doesn't seem like the place to reproduce a manual. What i was really objecting to is the notion that it's wrong to provide the answer, that somehow it's better to explain in each post how to read the manual. (ie. how to fish) In these cases especially, a quick answer solves their immediate problem and should provide inspiration for them on how to proceed in the future. Should, but doesn't necessarily. That's really up to the poster. I agree, but the statement you made was not leaving it up to the poster, you were leaving it up to your opinion, so that's where my response came from. Not sure what you mean here. Clearly? Yes, clearly. Oh sorry I missed have missed the memo. ;) lol, consider yourself CC'd IMHO, more people are served by seeing the correct answer. This is true. Wow, we agree on something ;o) Yeah I'm the kind of person that thinks a correct answer is more helpful than an incorrect answer. But that's just me. Then i'm not sure what we're disagreeing on in the first place. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: creating multi volume tar files .tar1, .tar2 etc
On Tue, 16 Sep 2003 18:31:32 -0700 Ian L [EMAIL PROTECTED] wrote: Hey all, i made a stupid assumption ... tar would do something intelligently. I was running a back up such as: tar -cvlf file.tar -L 50 /disk /disk2 What i thought this would do is back up /disk and /disk2 in multiple tar files of 500megs each. Instead, it creates file.tar and when it hits 500megs, waits for me to hit return and then just continues where it stopped at, except it starts file.tar over. Can someone tell me a way that i can tar up (tar, zip, compress i dont care) a group of directories, and have them split up by whatever size i set? and have it be multiple files ... i dont want to sit there and manually rename files or anything. I want to let it run and come back later and have X number of archive files. Ultimately these will be burned on a dvd drive on a windows machine. thanks, Hi Ian, tar -cvlO /disk /disk2 | split -b 50 - file.tar. The -O option to tar says to send the archive to the standard output. split will create files names file.tar. plus a two letter extension for as many files as it takes to store the input. To get a listing of what's in the archive, something like: cat file.tar.* | tar -tvf - Of course you could add the -z or -j to the tar command so that the output is compressed too. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: creating multi volume tar files .tar1, .tar2 etc
On Tue, 16 Sep 2003 19:05:31 -0700 Ian L [EMAIL PROTECTED] wrote: much thanks ... especially since my idea wont work since there's a 2gb file size limit as ed pointed out. Oh well ... i'll have a backup system working FWIW, The 2Gb limit has been removed in version 3 or better of NFS. The underlying filesystem has to support large files too, i use rieserfs and transfer 30+ GB files over NFS without a problem. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Latest iptables init scripts, and rmmod
On Tue, 16 Sep 2003 21:19:27 -0500 (EST) Mike Burger [EMAIL PROTECTED] wrote: Can anyone tell me whose bright idea it was to have the init script for iptables attempt to remove the iptables modules when one runs a service iptables stop? So far, it's caused one of my systems to crash and reboot, and another to lock up. Hmmm, sounds like a kernel bug. What kernel are you using? Have you upgraded to the latest RedHat supplied kernels? If it weren't for the fact that the output told me that it was trying to rmmod, I'd have nto known what to comment out of the init script. Good thing it was written well enough to tell you. ;o) Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Latest iptables init scripts, and rmmod
On Wed, 17 Sep 2003 12:58:22 +1000 Ian Mortimer [EMAIL PROTECTED] wrote: If it weren't for the fact that the output told me that it was trying to rmmod, I'd have nto known what to comment out of the init script. This is RedHat 8.0 right? The RedHat 9 init script doesn't do it. So far, it's caused one of my systems to crash and reboot, and another to lock up. Another side effect is that it causes hosts to hang during shut down or reboot at stopping iptables. A major pain if you're rebooting remotely. Might be a good idea to post it on bugzilla along with the fix. Be very surprised if this was always a problem with RH8 given its maturity. And the latest init scripts _do_ remove the module, they just use the modprobe command instead. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: 2.6.0-test5 won't boot
On Mon, 15 Sep 2003 13:56:07 +0800 Callan K L Tham [EMAIL PROTECTED] wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Saturday 13 September 2003 06:24, Kevin Breit wrote: Below is grub.conf: default=1 timeout=10 splashimage=(hd0,0)/grub/splash.xpm.gz title Red Hat Linux (2.6.0-test5) root (hd0,0) kernel /vmlinuz-2.6.0-test5 ro root=LABEL=/ You'll want to change root=LABEL=/ to root=/dev/hda3 or whatever holds your / partition. root=LABEL=/ is a RH thing...doesn't work with vanilla kernels. Hi Callan, It actually works without a problem with vanilla kernels. The LABEL feature is contained in the initrd code so as long as you supply a properly built initrd you'll be able to boot without having to hard code device paths. I have lots of problems getting test5 to compile though...maybe my skills aren't there yet. Recently a new menu option was added under Code Maturity Options: [*] Select only drivers expected to compile cleanly And everything should compile for you. If it doesn't, make sure to subscribe to the linux-kernel mailing list if you're not already. So many people on that list that know their stuff that you're bound to get a suggestion or complete solution quickly. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: the system activity message
On Sun, 14 Sep 2003 12:09:14 -0500 Dana Holland [EMAIL PROTECTED] wrote: I'm still getting these from cron: Invalid system activity file: /var/log/sa/sa13 I can't seem to find anything on google to help with this. Has anyone seen this before? Hi Dana, This file and the others in that directory are used to collect system activity information. You can safely delete them if they've become corrupt in some way and they will be regenerted. All you will lose is some historical data. If you decide you don't want to log this information anymore use rpm to remove the sysstat package. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Installing RH8 on proliant800
On Mon, 15 Sep 2003 02:34:21 -0700 (PDT) Nathalie Boulos [EMAIL PROTECTED] wrote: Hi everyone, I have a problem installing RH8 on a Compaq proliant 800. After I boot from CD, and whatever I put on the boot: command, the system freezes after writing the following: Running anaconda, the Red Hat Linux system installer - please wait... Probing for video card: Cirrus Logic GD543x Probing for monitor type: Unable to probe I have changed 3 monitors, 3 different brands, same result. is there any way not to probe for the monitor while installing? How can i bypass this problem? Hi Nathalie, One option is to use text mode for the install and work out your X windows issue afterward. After booting the install CD type linux text to use the text rather than graphical install. Good luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Installing RH8 on proliant800
On Mon, 15 Sep 2003 05:15:45 -0700 (PDT) Nathalie Boulos [EMAIL PROTECTED] wrote: Hello Sean, Well i've tried linux text and it did the same result. I tried the option lowres too, same thing. The installer tries to probe for the monitor first. How can i skip that probe? Nathalie, You can try adding noprobe and/or noddc onto the end of your boot line. Also, it may actually be failing after the last message you see in which case the problem may be with mouse detection. If you can plug a mouse (or different mouse) into the system you may get past the hang. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Installing RH8 on proliant800
On Mon, 15 Sep 2003 08:53:00 -0400 Sean Estabrooks [EMAIL PROTECTED] wrote: On Mon, 15 Sep 2003 05:15:45 -0700 (PDT) Nathalie Boulos [EMAIL PROTECTED] wrote: Hello Sean, Well i've tried linux text and it did the same result. I tried the option lowres too, same thing. The installer tries to probe for the monitor first. How can i skip that probe? Nathalie, You can try adding noprobe and/or noddc onto the end of your boot line. Also, it may actually be failing after the last message you see in which case the problem may be with mouse detection. If you can plug a mouse (or different mouse) into the system you may get past the hang. Nathalie, My memory was a bit off on this one the option is actually skipddc to skip the monitor probe. There is a list of all the boot options available to you on the RedHat site: http://redhat.pacific.net.au/redhat/linux/8.0/ja/doc/RH-DOCS/rhl-ig-x86-en-8.0/ch-bootopts.html Here is the boot line i'd try first: linux nousb skipddc noprobe apm=off Good luck, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Replace a string inside a file
On Mon, 15 Sep 2003 15:03:42 -0400 Keith Birchfield [EMAIL PROTECTED] wrote: Hello all, and thanks for previous help! :-) Can anyone give me a csh cheat to copy a file with a specific string replaced with another? For example copy foo1.txt to foo2.txt and replace each instance of blue with red sed -e 's/blue/red/g' foo1.txt foo2.txt Regards, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Network intallation bootdisk
On Mon, 15 Sep 2003 17:50:22 -0700 (PDT) truc nguyen [EMAIL PROTECTED] wrote: Now I want to create a network installation bootdisk to boot from the old computer and get boot.img from other computers through LAN. How do I create the network installation floppy ? Hi Truc, Check out the installation guide for Redhat 8 on their site, it has a section for making network boot disks: http://www.redhat.com/docs/manuals/linux/RHL-8.0-Manual/install-guide/s1-steps-install-cdrom.html#S2-STEPS-MAKE-DISKS Essentially what you'll need to do is make a floppy disk using bootnet.img. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: VERY slow NFS
On Sun, 14 Sep 2003 10:54:51 -0400 David Hart [EMAIL PROTECTED] wrote: Kernels on the server and all clients are configured for version 3. I cannot find the combination of server and client settings that will speed up NFS. Moving large files is considerably faster through FTP or moving them to the web root and using wget. Any suggestions? Hi David, You can try the rsize and wsize mount options mentioned in the man page for mount: rsize=8192,wsize=8192 This increases the block size for read and write operations respectively. It does increase the speed over what you get with the default values but i've never seen NFS as fast as ftp no matter how you tune it. Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: kernel panic after upgrade to 8.0
On Sun, 14 Sep 2003 10:52:53 -0500 Dana Holland [EMAIL PROTECTED] wrote: Apologies for cross posting, but I'm kind of in a bind here. I upgraded our webserver which was running 7.1 to 8.0 yesterday. The upgrade itself seemed to go fine, but then the machine wouldn't boot on it's own - it stuck in Loading linux. I did some searching and finally got a lilo.conf that seemed to load fine when I did /sbin/lilo. When I tried to reboot, it goes into kernel panic. Here's what I get: Mounting root filesystem mount: error 6 mounting ext3 pivotroot: prot_root (/sysroot,/sysroot/initrd) failed: 2 umount /initrd/proc failed: 2 kernel panic: No init found Try passing init: option to kernel I can still reboot just fine using the boot disk that was created when I did the upgrade. I'm heading back to google now to see what I can find. But if anyone has ideas, I would appreciate it. Hi Dana, The above error messages look like the root filesystem can not be found. What is the lilo.conf file you're using, and what drive/partition is your root filesystem on? Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: kernel panic after upgrade to 8.0
On Sun, 14 Sep 2003 11:12:34 -0500 Dana Holland [EMAIL PROTECTED] wrote: image=/boot/vmlinuz-2.4.20-20.8smp label=linux initrd=/boot/initrd-2.4.20-20.8smp.img read-only root=/dev/sdb2 Hi Dana, You didn't list any disks that match the root parameter above. Perhaps that should be root=/dev/sda2 ?? Sean. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: VERY slow NFS
On Sun, 14 Sep 2003 12:41:13 -0400 David Hart [EMAIL PROTECTED] wrote: On Sun, 2003-09-14 at 11:45, Sean Estabrooks wrote: Hi David, You can try the rsize and wsize mount options mentioned in the man page for mount: rsize=8192,wsize=8192 This increases the block size for read and write operations respectively. It does increase the speed over what you get with the default values but i've never seen NFS as fast as ftp no matter how you tune it. Hi: I've done that on the clients along with mountvers=3. How about on the server end? I have a feeling that the RH config is NOT consistent with version 3. Any suggestions? You can try adding the following lines to /etc/sysconfig/nfs: # 256kb recommended minimum size based on SPECsfs NFS benchmarks TUNE_QUEUE=yes NFSD_QS=524288 Which will double the default receive buffer size which is the only tunable i've fiddled with. Make sure that you're actually getting mounted with version 3 after the mount. Simply typing mount by itself will show you either a v2 or v3 beside the nfs mount. Also, you can try some transfers and use the nfsstat command to check out some statistics which might give you another clue. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: red hat linux 2.6.0-test5 build.
On Sat, 13 Sep 2003 12:06:35 +0530 srinivask [EMAIL PROTECTED] wrote: Hi, can any body sayy me if root dev is (3,2), what should be specified/appended in /etc/lilo.conf for 'root='. root=0302 or root=/dev/hda2 Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Source code for Install init
On 12 Sep 2003 00:00:14 -0400 Jason Dixon [EMAIL PROTECTED] wrote: Hey Jason, My apologies for the being unclear. I'm dealing with RHAS 2.1 (update 2). But you're right, it's in kernel-BOOT. It's the strangest thing, I *swear* I looked for that package before, and couldn't find it. :-P i think it's on the second CD so a quick check of the first wouldn't be enough ;o) P.S. Does the following sound like an initrd or vmlinuz error? looks more like an initrd problem. RAMDISK: Couldn't find valid RAM disk image starting at 0. Freeing initrd memory: 673k freed The initrd is never loaded. Either it is bad or can't be loaded from the media.. Perhaps even a bad boot parameter? EXT2-fs: unable to read superblock cramfs: wrong magic isofs_read_super: bread failed, dev=09:00, iso_blknum=16, block=32 Just sad thrashing about as it tries to boot without the benefit of a loaded ramdisk. Kernel panic: VFS: Unable to mount root fs on 09:00 Oh, looks like you are trying make a bootable tape, bare metal restore perhaps?I've never tried with Linux allthough i setup some HPUX systems to do this (Y2K, sigh).Not sure how you go about ensuring that the initrd is loaded from tape*.Would be interested to know how you make out with this. Cheers, Sean * initramfs in the 2.6 kernel is tailor made for this because it adds the image to the tail of the kernel itself so you don't have to load it from a boot partition. The initramfs tools are still imature though. -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Source code for Install init
On 12 Sep 2003 10:07:43 -0400 Jason Dixon [EMAIL PROTECTED] wrote: Kernel panic: VFS: Unable to mount root fs on 09:00 Oh, looks like you are trying make a bootable tape, bare metal restore perhaps?I've never tried with Linux allthough i setup some HPUX systems to do this (Y2K, sigh).Not sure how you go about ensuring that the initrd is loaded from tape*.Would be interested to know how you make out with this. Close, but not cigar. I'm actually attempting to modify init.c in the ramdisk for a custom network kickstart. It's actually nothing special, just a timely sleep(30). But I've had lots of problems with the Enterprise spec... particularly with the helpfile sections in %build. They keep crashing on one part of another. Commenting them out (hell, they're not needed anyways) allows the build to complete, but we're obviously left with an unusable ramdisk. Well that's what i get for guessing before my first coffee of the day. ;o) Perhaps my eyes are still crossed but it appears to be trying to use the tape device (09:00) to find the root fs. Maybe an rdev or a root= parameter is needed ? Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: 2.6.0-test5 won't boot
On Fri, 12 Sep 2003 15:36:20 -0400 Kevin Breit [EMAIL PROTECTED] wrote: Hi Kevin, I compiled 2.6.0-test5 (latest) today and am having a hard time booting it. During boot, I get: Kernel panic. No kernel found. Trey passing init= to kernel. The message you're seeing is actually: No init found. Try passing init= option to kernel. This means that it was unable to mount your root filesystem. Would need more details to know why. My .img file is specified in grub.conf and does exist in the proper directory. As does the kernel. Since you're compiling your own kernel don't use modules. They just make life more difficult and in fact make your kernel slightly slower. If you compile just the options you need into one static kernel you don't need an initrd .img file at all. If you have some special need to use modules make sure you upgrade to the latest modutils, the 2.6 kernel requires updated module handing programs. Cheers, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: root password and su (maybe)
On Thu, 11 Sep 2003 12:02:59 +0100 Kelerion [EMAIL PROTECTED] wrote: I was wondering if there was another way around this.. I was thinking that there might be a way to setup su root to accept *my* login as a trusted user and therefore not ask me for a password when I su root.. then I can simply su and change the password back again that way.. One option is to create a sudo entry for yourself to allow use of the passwd command to change root password. If you really want to go overboard you can create a second root account: useradd -u 0 -o noboss ; passwd noboss Securely yours, Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
Re: Moving Viewport with keyboard?
On Thu, 11 Sep 2003 12:31:07 +0200 T. Ribbrock [EMAIL PROTECTED] wrote: Hi all, I have a laptop with a resolution of 800x480. I've set the virtual resolution to 800x600. It all works just dandy, but I have one question: Is there any way to scroll around the screen without having to use the mouse? If you're using Gnome you could use the keyboard-mouse provided by the accessibility features: run gnome-accessibility-keyboard-properties and enable both Enable keyboard accessibility features and Enable Mouse Keys After this you can use the numeric keypad to move the mouse pointer around and portions the virtual screen into view. Sean -- redhat-list mailing list unsubscribe mailto:[EMAIL PROTECTED] https://www.redhat.com/mailman/listinfo/redhat-list
