Re: Frees/wan setup problems
On Wed, 2003-02-26 at 17:16, Kenneth E. Lussier wrote: On Wed, 2003-02-26 at 15:03, Cole Tuininga wrote: But not via the VPN: traceroute 192.168.2.1 traceroute to 192.168.2.1 (192.168.2.1), 30 hops max, 38 byte packets 1 * * * (etc) What about from the subnet to the gateway? I'm not sure if you mean the closer gateway or the further one, but from 192.168.1.68: $ traceroute 192.168.1.1 traceroute to 192.168.1.1 (192.168.1.1), 30 hops max, 38 byte packets 1 192.168.1.1 (192.168.1.1) 0.445 ms 0.194 ms 0.173 ms $ traceroute 192.168.2.1 traceroute to 192.168.2.1 (192.168.2.1), 30 hops max, 38 byte packets 1 192.168.1.1 (192.168.1.1) 0.340 ms 0.181 ms 0.161 ms 2 * * * (*sputter* *cough* *choke*) 2) 192.168.1.0/24-to-opposite gateway 3) 192.168.2.0/24-to-opposite gateway You're basically saying to just add these two in? Yes. Ok - so here's what the appropriate sections of my ipsec.conf look like now: conn panam-cole-ss left = 63.127.199.26 leftsubnet = 192.168.2.0/24 leftnexthop = 63.127.199.25 # RSA 2192 bits inside Sat Feb 22 04:28:58 2003 leftrsasigkey=0sAQNkta3s7v3B [snip] right = 209.187.117.100 rightsubnet = 192.168.1.0/24 rightnexthop = 209.187.117.65 # RSA 2192 bits deb-box Tue Feb 25 13:13:11 2003 rightrsasigkey=0sAQPBb4MhWjn [snip] auto = start conn panam-cole-gg left = 63.127.199.26 leftnexthop = 63.127.199.25 # RSA 2192 bits inside Sat Feb 22 04:28:58 2003 leftrsasigkey=0sAQNkta3s7v3B [snip] right = 209.187.117.100 rightnexthop = 209.187.117.65 # RSA 2192 bits deb-box Tue Feb 25 13:13:11 2003 rightrsasigkey=0sAQPBb4MhWjn [snip] auto = start conn panam-cole-gs left = 63.127.199.26 leftnexthop = 63.127.199.25 # RSA 2192 bits inside Sat Feb 22 04:28:58 2003 leftrsasigkey=0sAQNkta3s7v3Bv [snip] right = 209.187.117.100 rightsubnet = 192.168.1.0/24 rightnexthop = 209.187.117.65 # RSA 2192 bits deb-box Tue Feb 25 13:13:11 2003 rightrsasigkey=0sAQPBb4MhWjnp9 [snip] auto = start conn panam-cole-sg left = 63.127.199.26 leftsubnet = 192.168.2.0/24 leftnexthop = 63.127.199.25 # RSA 2192 bits inside Sat Feb 22 04:28:58 2003 leftrsasigkey=0sAQNkta3s7v3B [snip] right = 209.187.117.100 rightnexthop = 209.187.117.65 # RSA 2192 bits deb-box Tue Feb 25 13:13:11 2003 rightrsasigkey=0sAQPBb4MhWjnp9Y4 [snip] auto = start After setting these up and activating them, initially, my ssh connection to the remote gateway (63.127.199.26) stalls for a minute or two and then comes back. Routing tables look like this: 192.168.1.1: # route -n Kernel IP routing table Destination Gateway Genmask Flags Metric RefUse Iface 63.127.199.26 209.187.117.65 255.255.255.255 UGH 0 00 ipsec0 209.187.117.64 0.0.0.0 255.255.255.192 U 0 00 eth1 209.187.117.64 0.0.0.0 255.255.255.192 U 0 00 ipsec0 192.168.2.0 209.187.117.65 255.255.255.0 UG0 00 ipsec0 192.168.1.0 0.0.0.0 255.255.255.0 U 0 00 eth0 0.0.0.0 209.187.117.65 0.0.0.0 UG0 00 eth1 192.168.2.1: # route -n Kernel IP routing table Destination Gateway Genmask Flags Metric RefUse Iface 209.187.117.100 63.127.199.25 255.255.255.255 UGH 0 00 ipsec0 63.127.199.24 0.0.0.0 255.255.255.252 U 0 00 eth0 63.127.199.24 0.0.0.0 255.255.255.252 U 0 00 ipsec0 192.168.2.0 0.0.0.0 255.255.255.0 U 0 00 eth1 192.168.1.0 63.127.199.25 255.255.255.0 UG0 00 ipsec0 0.0.0.0 63.127.199.25 0.0.0.0 UG0 00 eth0 You can try forcing a route between the two gateways, which might help. Can you send me the output of a pluto barf? (BTW... Who *NAMES* these things??? ;-) A barf from when? Startup? When I try to send ICMP packets through? TCP packets? C. All of the above. ;-) I'll generate this in a little bit and mail it to Kenny off list. Anybody else masochistic ... er ... interested enough in looking at the logs can email me and I'll be happy to send a copy. It looks like the requests are going out, but the response is not coming back. You may want to run tcpdump without specifying an interface to see even more of the traffic. From this it looks like the pings are going out to the box but are not routed back properly. Ok, this time around, I'm doing a: # tcpdump -i any | grep icmp From 192.168.1.68, I attempt to ping 192.168.1.69. Here are the tcpdump results: 08:55:36.694803 192.168.1.68 192.168.2.69: icmp: echo request (DF) 08:55:36.694985 192.168.1.68 192.168.2.69: icmp: echo request (DF) 08:55:36.748865 192.168.2.69 192.168.1.68: icmp: echo reply 08:55:37.687780
Future meeting topics
Here's a few ideas, some of them relevant to what I'm going through and ideas from last night's meeting: Debugging kernel oopses (there's a lot of slides on the topic, but there they reference what the speaker is saying) LVM on single-drive machines (install, use, advantages, etc.) Writing presentations using mgp (a few of us use it, including me) Bluetooth / IEEE1394 use under Linux For the 'newbies': New features in RH 8.1 Comparison of programming languages available on Linux Book reviews might be an idea, if even just on the mailing list BTW, my slides from last night's presentation are available at http://www.wayga.org/melba/02_2003 -Mark pgp0.pgp Description: PGP signature
Re: Linux booting problems
[EMAIL PROTECTED] wrote: In a message dated: Thu, 27 Feb 2003 15:30:58 EST Kenny Donahue said: ext3 comes up then it try's to load the ext3 module. [nit-picking: tries not try's :) ] it fails with undefined symbols in jdb.o and ext3.o. Why is it trying to load modules that are suposed to be in the kernel I looked in /lib/modules/linux-2.4.7-10custom/ and these modules DONT EXIST!!! Chances are there's a modules.conf or something which loaded these from the default kernel, and it's still trying... make modules_install but you probably just blew them away with this step here. (of course, I could be wrong :) make install I recommend not doing the make install, rather, manually copying arch/arch/boot/bzImage to /boot/vmlinuz-whatever and editing the lilo.conf to add a new kernel. Then run lilo to recognize this new kernel and boot from it. That way you don't lose your old kernel. I did do a make modules. Sorry about the spelling but I'm typing in anger. I save the output from my make install and it does A LOT more then just copy hte bzImage vmlinuz and System map make install /tmp/makeInstallOutput Thanks, Kenny ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
On Thu, Feb 27, 2003 at 03:50:49PM -0500, Kenny Donahue wrote: [EMAIL PROTECTED] wrote: I recommend not doing the make install, rather, manually copying arch/arch/boot/bzImage to /boot/vmlinuz-whatever and editing the lilo.conf to add a new kernel. Then run lilo to recognize this new kernel and boot from it. That way you don't lose your old kernel. I did do a make modules. Sorry about the spelling but I'm typing in anger. I save the output from my make install and it does A LOT more then just copy hte bzImage vmlinuz and System map make install /tmp/makeInstallOutput IIRC, the make install also fires off lilo, or at least configures lilo for you. I'd stick to the previously-given advice and move the bzImage and System map manually into /boot. I spent the first few days this week wrestling with the kernel, so this is a topical discussion for me. -Mark pgp0.pgp Description: PGP signature
Re: Linux booting problems
Mark Komarinski wrote: On Thu, Feb 27, 2003 at 03:50:49PM -0500, Kenny Donahue wrote: [EMAIL PROTECTED] wrote: I recommend not doing the make install, rather, manually copying arch/arch/boot/bzImage to /boot/vmlinuz-whatever and editing the lilo.conf to add a new kernel. Then run lilo to recognize this new kernel and boot from it. That way you don't lose your old kernel. I did do a make modules. Sorry about the spelling but I'm typing in anger. I save the output from my make install and it does A LOT more then just copy hte bzImage vmlinuz and System map make install /tmp/makeInstallOutput IIRC, the make install also fires off lilo, or at least configures lilo for you. I'd stick to the previously-given advice and move the bzImage and System map manually into /boot. I spent the first few days this week wrestling with the kernel, so this is a topical discussion for me. -Mark The make install claims that it modifies lilo.conf but it doesn't. I modify it myself. I am recompiling now (again :( ) . I will try it by hand. My main questions was why was it trying to load modules when I built it into the kernel? I looked at modules.conf. that's ok. I will look in modules.dep after I'm done. One more quick question do I need jbd with ext3? I am not including it right now. Thanks, Kenny ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
In a message dated: Thu, 27 Feb 2003 16:09:50 EST mike ledoux said: Yes. ext3 depends on jbd. If you didn't compile in jbd, that explains why it is looking for the module; if you didn't build the module, that explains the error. I sit corrected :) It's been a while since I looked at a kernel config and needed ext3 (I've been using XFS and reiserfs). -- Seeya, Paul -- Key fingerprint = 1660 FECC 5D21 D286 F853 E808 BB07 9239 53F1 28EE It may look like I'm just sitting here doing nothing, but I'm really actively waiting for all my problems to go away. If you're not having fun, you're not doing it right! ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
In a message dated: Thu, 27 Feb 2003 16:05:26 EST Kenny Donahue said: The make install claims that it modifies lilo.conf but it doesn't. It's for precisely this reason I stopped using 'make install'. My main questions was why was it trying to load modules when I built it into the kernel? Stupid question, but are you sure you were booting the right kernel? Could you have not changed the default kernel label to point to yours, or forgotten to specify at boot time to use the new kernel? As I'm sure you know, just running lilo to recognize your new kernel does no good if there's a default label pointing to some *other* kernel. So, if the default tried to install the ext3 mod, and it's no longer there, you're screwed. One more quick question do I need jbd with ext3? I am not including it right now. I don't think so. I don't ever remember using it. -- Seeya, Paul -- Key fingerprint = 1660 FECC 5D21 D286 F853 E808 BB07 9239 53F1 28EE It may look like I'm just sitting here doing nothing, but I'm really actively waiting for all my problems to go away. If you're not having fun, you're not doing it right! ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
mike ledoux wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Thu, Feb 27, 2003 at 04:05:26PM -0500, Kenny Donahue wrote: My main questions was why was it trying to load modules when I built it into the kernel? I looked at modules.conf. that's ok. I will look in modules.dep after I'm done. One more quick question do I need jbd with ext3? I am not including it right now. Yes. ext3 depends on jbd. If you didn't compile in jbd, that explains why it is looking for the module; if you didn't build the module, that explains the error. so. why is ext3 on by default but jbd off by default? The only why I figured this out was by looking at the modules.dep file. Thanks, Kenny ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
At the risk of sounding like a smart aleck, I would very strongly suggest that anyone building a test kernel like this on an i386-based system, do so by doing a 'make bzdisk' and have it expectorate the kernel under test onto a floppy. This will help reduce the finger-pointing around LILO and /boot, etc., and therefore reduce the amount of troubleshooting necessary, and allow you to go back to a known state to rebuild your kernel again. Also, RH 7.2 is getting a bit old, and it's possible (and I honestly don't know for sure, since I run SuSE) that there may be some subtle issues around the compiler suite in 7.2. (Outdated libraries or some such). I'm sure that others may be able to tell me why I'm wrong about this, however :-) HTH, Bayard ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
'make bzSomethingOtherThanOnAFloppy [was Re: Linux booting problems ]
In a message dated: Thu, 27 Feb 2003 16:33:01 EST Bayard R. Coolidge said: At the risk of sounding like a smart aleck, I would very strongly suggest that anyone building a test kernel like this on an i386-based system, do so by doing a 'make bzdisk' and have it expectorate the kernel under test onto a floppy. Here's a stupid question. Dell recently announced that they're getting rid of floppy drives in their systems. I expect most other manufacturers to eventually follow suit. So, when this happens, and we're all living in a floppy-less world, how do then do things like 'make bzdisk' ? -- Seeya, Paul -- Key fingerprint = 1660 FECC 5D21 D286 F853 E808 BB07 9239 53F1 28EE It may look like I'm just sitting here doing nothing, but I'm really actively waiting for all my problems to go away. If you're not having fun, you're not doing it right! ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: 'make bzSomethingOtherThanOnAFloppy [was Re: Linux booting problems]
So, when this happens, and we're all living in a floppy-less world, how do then do things like 'make bzdisk' ? Beats me, although to be honest, I've had the same problem on Alpha for a couple of years, although it's been months since I've booted my Alpha, much less built a kernel for it. Clearly, it will be a bit trickier to do experimental kernel builds. The thing that caught my attention in the ongoing discussion was the automatic invocation of LILO. I was not aware that that was the case, but since it's been a couple of years since I've built a kernel that wasn't on a floppy, I guess I'm just ignorant. I suspect that we'll all have to get used to being very, very careful about what /boot/System.map points to, and what kernels are called what in /boot and so forth. I haven't tried booting a SCSI ZIP disk, but that might be one possibility IMNSHO. B. ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Eric Raymond's writings.
Has anyone noticed that the site that once contained Eric Raymond's writings: www.tuxedo.org/~esr/ is no longer accessable via the web. The link is greeted with a script that redirects one to a selection of random sites dealing with Open Source. Although Mr. Raymonds writings were always treated by me as a gift, I somehow feel robbed. The random links are a treasure of information however I miss the freedom of reading one of Open Source's premier advocates. This is not a technical issue, just an issue that makes one say hmm. David -- ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Kenny Donahue [EMAIL PROTECTED] writes: Ok deeep breath. I'm ready to though Linux out. I've been fighting with a Linux kernel build for a while now. I'm getting very angry. I can load a Redhat 7.2 distribution and everything works. I need to patch the kernel for a project at work. so, I do my patch (../fs/proc/base.c if it makes a difference) then [snip] I've been running Redhat for ages, and I've found that it's easier to stick with kernel rpms. Fortunately the stock kernel sources include an rpm target, so building rpms from the virgin kernel source should be fairly straightforward. The trick is to load the same config that your system is already using, which you'll find in one of the files under /usr/src/linux-2.4/configs, while in make xconfig. Then apply your changes from that known-good setting. Try this: make mrproper make xconfig load config from /usr/src/linux-2.4/configs/kernel-2.4.18-i686.conf ig add your scsi driver into the kernel make dep make rpm sudo rpm -i /usr/src/redhat/RPMS/i686/kernel-2.4.18.i686.rpm edit lilo.conf ... When I do this, I like to do the full build without adding my patches, so I can explicitly verify that the pre-patched setup is working. Note that you want to install (-i) the new kernel, not upgrade (-U). That way you don't lose the older kernel that already works correctly, so if the new kernel fails you can still boot the system. - -- John Abreau / Executive Director, Boston Linux Unix Email [EMAIL PROTECTED] / WWW http://www.abreau.net / PGP-Key-ID 0xD5C7B5D9 PGP-Key-Fingerprint 72 FB 39 4F 3C 3B D6 5B E0 C8 5A 6E F1 2C BE 99 -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.1 (GNU/Linux) Comment: Exmh version 2.6 02/09/2003 iQCVAwUBPl6dglV9A5rVx7XZAQLnoQP+L2InebdmlniwFBwpnK+GtjGqzikHthyL sjLSMaGgiCWKNwGm3GlSAR61UsmLB+Ye5HHiZ4U8WFgPrh0QdBSRZweqMdvqvTQH UxRDTnJwU8Oph73bxAgikGFa+O06/veeEN/wYRm8Z814iWvApPKbnWXqmrJo7jWK n7Hhs/YUsEc= =Pq/5 -END PGP SIGNATURE- ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Eric Raymond's writings.
David Richter [EMAIL PROTECTED] writes: Has anyone noticed that the site that once contained Eric Raymond's writings: www.tuxedo.org/~esr/ is no longer accessable via the web. The link is greeted with a script that redirects one to a selection of random sites dealing with Open Source. I noticed this a while ago too. I don't know what the reason is for the random redirections (if anybody knows, please clue me in). But anyways, here's what seems to be a reliable home page: http://catb.org/~esr/ Regards, --kevin -- It is seldom that liberty of any kind is lost all at once. -Hume ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: Linux booting problems
It's been a while since I've played with RedHat, but IIRC, when using a stock kernel, you have to uncomment the line in the make file that points the image to be placed in /boot... -- Things are fine, the upcoming semester approaches like a brick wall and we're in a 1962 Corvair with no brakes. - Paul Sand Cole Tuininga Lead Developer Code Energy, Inc [EMAIL PROTECTED] PGP Key ID: 0x43E5755D ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: 'make bzSomethingOtherThanOnAFloppy [was Re: Linux booting problems]
I use grub and I remember from using LILO that it is very easy to set up the menu to select multiple kernel versions. So, you build your experimental kernel, store it next to your main kernel, add a lilo.conf or grub.conf entry for your new kernel, then reboot and choose the experimental kernel. I've been doing this for several years. I've never bothered with make bzdisk. BTW, my main workstation has not had a floppy drive for over two years now. I built it with one, but took it out because I was never using it. Bayard R. Coolidge wrote: So, when this happens, and we're all living in a floppy-less world, how do then do things like 'make bzdisk' ? Beats me, although to be honest, I've had the same problem on Alpha for a couple of years, although it's been months since I've booted my Alpha, much less built a kernel for it. Clearly, it will be a bit trickier to do experimental kernel builds. The thing that caught my attention in the ongoing discussion was the automatic invocation of LILO. I was not aware that that was the case, but since it's been a couple of years since I've built a kernel that wasn't on a floppy, I guess I'm just ignorant. I suspect that we'll all have to get used to being very, very careful about what /boot/System.map points to, and what kernels are called what in /boot and so forth. I haven't tried booting a SCSI ZIP disk, but that might be one possibility IMNSHO. B. ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Colors INSIDE of windows?
Bill Sconce asks about things like application color choices. Bill, Traditionally, these things were X resources. There's a whole boodle of stuff somewhere in the X documntation about class name versus item name, etc., and how to wildcard things so that you can have all xterms do this, except that a certain xterm will do that, or something of that ilk. I can remember finding this stuff in the man pages for the base X system, but you had to dig for it, and it was hard to tell what they were saying unless you already knew what you were looking for. The resources are maintained in the X server, which is nice, since apps that you run remotely with, say, xon will ask your server for the settings that you prefer on that workstation. xrdb is the program that loads your .Xdefaults into the server. It has to get run when you start the server or log in to xdm, so it gets invoked in one of your personal .x startup files, like .xclient, or .xinitrc, or if you don't have a personal one, various vintages of the system-wide startup files might try to look for ~/.Xdefaults for you. I gave up trying to follow the improvements before desktops replaced window managers, so the exact path of the data is a matter for exploration. Things like backgroundcolor are common enough, but not every program documents their resource names well (at least in the online documents that I've found). It can be a chore to find the resource name that will do what you want. In the past I've found stuff in the application's man or info pages. Of course, the developer of any particular application is free to ignore the resource system. And some apps may have ways of overriding the resource selections. For example, emacs will follow the resource selections, but you can also customize that stuff via the .emacs file, the emacs internal customize system, and the command line. Which one wins when something gets specified in multiple palces is up to the developer, but it usually isn't the X resources. And if you're running Ximian, or one of the other modern environments, who knows what they may have done to avoid giving the user the dredful pain of having a customizable system. Bill ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
Re: 'make bzSomethingOtherThanOnAFloppy [was Re: Linux bootingproblems ]
On Thu, 27 Feb 2003, at 4:45pm, [EMAIL PROTECTED] wrote: Here's a stupid question. Dell recently announced that they're getting rid of floppy drives in their systems. Just FYI, what they actually announced was that they are making the floppy drive an option on their Precision Workstation line. Of course, you can expect other models and vendors to follow eventually. So, when this happens, and we're all living in a floppy-less world, how do then do things like 'make bzdisk' ? I haven't used make bzdisk in ages. Floppies are so damn slow. When I build a new kernel for test, I drop it into /boot/ (along with any needed support files), and define a new GRUB entry, GRUB makes it very easy to switch kernels, or modify kernel options at boot. (Before GRUB, I used LILO, which was more restrictive but worked well enough.) And I always kept my stock kernel around, to boot in case of emergency. If I somehow manage to get the system so hosed my backup kernels don't work (in which case, I'd probably be hosed even if I *had* used bzdisk), I boot from CD and start repairing the system that way. If you really wanted some kind of functionality like bzdisk, I suppose you could script something to automatically create a bootable CD-RW, but why bother? :) -- Ben Scott [EMAIL PROTECTED] | The opinions expressed in this message are those of the author and do | | not represent the views or policy of any other person or organization. | | All information is provided without warranty of any kind. | ___ gnhlug-discuss mailing list [EMAIL PROTECTED] http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss
