Re: VDISK Swap space on RHAS 3
Reuscher, Robert A [ITS] wrote: I'm having some minor issues with VDISK Swap space on Redhat AS 3 (Taroon Update 2) and the dasd_diag_mod driver. It's 31bit mode, so that should work (from what I understand). I'm using SWAPGEN (V5) for format the VDISK swap space prior to Linux IPL, that's working fine. However when RHAS detects the disk it loads the FBA driver, not the diag driver. How can I get it to use the diag driver, since it supposed to better I don't have access to a current Red Hat AS, but I do recall there was a bug in the dasd driver that called the discipline handlers in the wrong order. The trick I used there when the FBA driver picked it up: echo set device 200 off /proc/dasd/devices modprobe dasd_diag_mod Upon loading, the diag driver also looks around for disks it could use, and it will enable the disk that you just removed from the fba driver. Later patches assured that the diag driver was tried first. Rob -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Tcpdump on Linux-390.
Vic Cross wrote: On Friday 20 August 2004 07:09, Rob van der Heij wrote: Rv Yep, SuSE had some patches on top that know about the unique aspects of Rv the OSA devices. It was actually a wrapper script written by someone (an IBMer?) to redirect stdin and/or stdout of the tcpdump program and add a fake LLC header to the packets captured off the network. SuSE included the wrapper script in their distro (from SLES8, IIRC). I know better than argue with you... I recall the early version had tweaked one of the formats since I could not use the same thing to decode a tcpdump stream from my PC anymore. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Tcpdump on Linux-390.
On Friday 20 August 2004 18:23, Rob van der Heij wrote: Rv I know better than argue with you... I recall the early version had Rv tweaked one of the formats since I could not use the same thing to Rv decode a tcpdump stream from my PC anymore. Good point. If you run tcpdump-qeth -r on a file you captured on a PC, I suppose you will get garbage because the script will add an extra LLC header. The real tcpdump program should have been used in this case. Hopefully SuSE's package did not make tcpdump a link to tcpdump-qeth... :) Last year I was using tcpdump-qeth to capture a trace (using the -w option) from a Linux guest on a VSWITCH, and I was able to view the result with standard ethereal on my laptop. So I know it works! ;) (For the ultra-curious, on page 36 of the VSWITCH Redpaper you can see the result of such a capture. Notice the all-zeroes MAC addresses in the Ethernet header -- that 'empty' Ethernet-II header is what tcpdump-qeth adds to the packet flow to keep tcpdump and friends happy.) Cheers, Vic PS: The same Redbook credits tcpdump-qeth to Holger Smolinski. See, I knew it was an IBMer! Sorry, Holger ;) -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Dynamic I/O Configuration and removing eth1
On Friday 20 August 2004 09:07, Peter E. Abresch Jr. - at Pepco wrote: PE Thanks for responding. Here is what I get: PE PE linuxd01:/etc/sysconfig/network # lsmod PE Module Size Used byNot tainted PE qdio 37040 0 PE 8021q 15256 0 (unused) PE nfsd 80392 4 (autoclean) PE ipv6 329288 -1 (autoclean) PE key41840 0 (autoclean) [ipv6] PE lcs29248 1 PE dummy0 1108 1 PE ctc49824 3 PE fsm 2032 0 [ctc] PE lvm-mod70408 11 (autoclean) PE dasd_eckd_mod 56648 5 PE dasd_mod 49748 7 [dasd_eckd_mod] PE reiserfs 254940 7 PE PE What do I removed. I still have eth0 going so I cannot remove lcs. Firstly, make sure you've removed the old LCS definition for eth1 from /etc/chandev.conf. Then, two things to try: echo reset_conf /proc/chandev echo read_conf /proc/chandev echo reprobe /proc/chandev This *might* cause the LCS driver to release its definition for eth1. Otherwise, change the qeth1 at the start of the line for your new interface in /etc/chandev.conf to something else -- I'd say qeth-1, to let it pick -- and repeat the above three commands. Your interface will probably be eth2 until your next reboot, however, so this might be more trouble to you than it is worth to do it dynamically... Cheers, Vic Cross -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
cpint-1.1.6 make errors on RHEL 3.0
Hi, I am trying to 'make' cpint-1.1.6 downloaded from linuxvm.org/patches on my RHEL 3.0 Update 2 system. Here is what I receive when I run 'make' . Looking for help on what modifications need to be made to make it to work. [EMAIL PROTECTED] cpint-1.1.6]# make gcc -D__KERNEL__ -DMODULE -Wall -O2 -I/lib/modules/`uname -r`/build/include -I/usr/include -c -o main.o main.c In file included from main.c:64: /usr/include/linux/proc_fs.h:70: syntax error before atomic_t /usr/include/linux/proc_fs.h:73: syntax error before '}' token /usr/include/linux/proc_fs.h:203: syntax error before void In file included from /usr/include/linux/sched.h:14, from /usr/include/asm-s390/uaccess.h:17, from /usr/include/asm/uaccess.h:4, from main.c:70: . (not including the rest of the messages to make it more readable) Regards, Sandeep Batta Host System Services, Phone : 919-224-1282 email : [EMAIL PROTECTED] -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Dynamic I/O Configuration and removing eth1
On Thu, 2004-08-19 at 17:58, Peter E. Abresch Jr. - at Pepco wrote: I assume this is cause by eth1 still being defined. How do I delete eth1? In /etc/modules.conf: alias eth1off Then remove|rename /etc/sysconfig/network/ifcfg-eth1. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Tcpdump on Linux-390.
I am on vacation today so I will try tcpdump-qeth on Monday. I am using qdio qeth on a guest LAN to an OSA card. I am only interested in traffic to and from this vmid. I seem to get either everything on the LAN but with an unknown protocol or, if I specify an expression like host host-name, I get nothing. _/) Tom Shilson ~GEDW VM System Services Aloha Tel: 651-733-7591 tshilson at mmm dot com Fax: 651-736-7689 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: VDISK Swap space on RHAS 3
I do recall there was a bug in the dasd driver that called the discipline handlers in the wrong order. We did this with SuSE during the Domino residency last year and I thought we did this with Red Hat too, but don't see it documented in the redbook (SG24-7021). It is documented for SuSE on p 125 that you have to add the module to /etc/sysconfig/kernel and run mkinitrd. RHEL-3 does not seem to read from /etc/sysconfig/kernel and the syntax for mkinitrd is different from SuSE's. I believe you could go to /boot, back up your good initrd and do something like: mkinitrd --preload jbd --preload ext3 --preload dasd_diag_mod initrd-2.4.21-4.EL.img `uname -r` Then if you create the vdisk swap using SWAPGEN with the (DIAG option and set up the swap space in fstab, I believe Linux will pick up the right module. I am unable to test fully at this time. Hope this helps. -Mike MacIsaac, IBM [EMAIL PROTECTED] (845) 433-7061 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: Dynamic I/O Configuration and removing eth1
Thanks for the reply. I discovered my problem. It came to me in a dream in the form of a beautiful girl. Too bad she was a penguin. Anyway, I digress. I have 4 OSAs installed but are only currently using two on this particular LPAR. I did not have noauto: in my /etc/chandev.conf, and it appears it was picking up all my OSAs even though I had not configured them with ifconfig. I changed qeth1 to qeth3 in /etc/chandev.conf and it works great. Peter Vic Cross [EMAIL PROTECTED] auTo Sent by: Linux on [EMAIL PROTECTED] 390 Port cc [EMAIL PROTECTED] IST.EDU Subject Re: Dynamic I/O Configuration and removing eth1 08/20/2004 05:37 AM Please respond to Linux on 390 Port [EMAIL PROTECTED] IST.EDU On Friday 20 August 2004 09:07, Peter E. Abresch Jr. - at Pepco wrote: PE Thanks for responding. Here is what I get: PE PE linuxd01:/etc/sysconfig/network # lsmod PE Module Size Used byNot tainted PE qdio 37040 0 PE 8021q 15256 0 (unused) PE nfsd 80392 4 (autoclean) PE ipv6 329288 -1 (autoclean) PE key41840 0 (autoclean) [ipv6] PE lcs29248 1 PE dummy0 1108 1 PE ctc49824 3 PE fsm 2032 0 [ctc] PE lvm-mod70408 11 (autoclean) PE dasd_eckd_mod 56648 5 PE dasd_mod 49748 7 [dasd_eckd_mod] PE reiserfs 254940 7 PE PE What do I removed. I still have eth0 going so I cannot remove lcs. Firstly, make sure you've removed the old LCS definition for eth1 from /etc/chandev.conf. Then, two things to try: echo reset_conf /proc/chandev echo read_conf /proc/chandev echo reprobe /proc/chandev This *might* cause the LCS driver to release its definition for eth1. Otherwise, change the qeth1 at the start of the line for your new interface in /etc/chandev.conf to something else -- I'd say qeth-1, to let it pick -- and repeat the above three commands. Your interface will probably be eth2 until your next reboot, however, so this might be more trouble to you than it is worth to do it dynamically... Cheers, Vic Cross -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 This Email message and any attachment may contain information that is proprietary, legally privileged, confidential and/or subject to copyright belonging to Pepco Holdings, Inc. or its affiliates (PHI). This Email is intended solely for the use of the person(s) to which it is addressed. If you are not an intended recipient, or the employee or agent responsible for delivery of this Email to the intended recipient(s), you are hereby notified that any dissemination, distribution or copying of this Email is strictly prohibited. If you have received this message in error, please immediately notify the sender and permanently delete this Email and any copies. PHI policies expressly prohibit employees from making defamatory or offensive statements and infringing any copyright or any other legal right by Email communication. PHI will not accept any liability in respect of such communications. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: [Linux/390] Re: Tcpdump on Linux-390.
If you're using a 2.6.x kernel, you can turn on LLC emulation via proc for the device (It's in the channel howto stuff for 2.6, I can find it later). This lets an unmodified tcpdump (and other sniffing tools) work. -m On Fri, Aug 20, 2004 at 07:49:16PM +0200, Rob van der Heij wrote: Tom Shilson wrote: I am on vacation today so I will try tcpdump-qeth on Monday. I am using qdio qeth on a guest LAN to an OSA card. I am only interested in traffic to and from this vmid. Good for you! I tend to have time to play with those things only when I am on vacation. I have been talking to folks in z/VM development about a (properly secured) way to see more than just your own traffic, but that will probably wait until someone there has vacation ;-) So tcpdump will only show you your own traffic, which is what you wanted. I seem to get either everything on the LAN but with an unknown protocol or, if I specify an expression like host host-name, I get nothing. I just checked that the tcpdump that comes with SuSE SLES8 works out-of-the-box for qeth. There is no wrapper thing included, so I believe this was indeed modified to match the missing link header. Rob -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390 -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390
Re: [Linux/390] Re: Tcpdump on Linux-390.
Unfortunately this is the 2.4 kernel. Thanks anyway. _/) Tom Shilson ~GEDW VM System Services Aloha Tel: 651-733-7591 tshilson at mmm dot com Fax: 651-736-7689 Linux on 390 Port [EMAIL PROTECTED] wrote on 08/20/2004 12:52:24 PM: If you're using a 2.6.x kernel, you can turn on LLC emulation via proc for the device (It's in the channel howto stuff for 2.6, I can find it later). This lets an unmodified tcpdump (and other sniffing tools) work. -- For LINUX-390 subscribe / signoff / archive access instructions, send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit http://www.marist.edu/htbin/wlvindex?LINUX-390