VLAN support in SuSE zLinux
Hi, We're using the 802.1q VLAN support in SuSE zLinux over a VSWITCH. Works great. Yesterday, I tried to set up another guest with the same VLAN support (same VLAN, too). The packets exit the vlan interface and hit the base interface, then exit to the VSWITCH. Looks fine. Packets then arrive from the VSWITCH to the base interface, but they never make it to the vlan interface. I am pretty sure that I have set everything up to use vlan-tagged frames only. The VSWITCH is set up correctly, and the guest has appropriate GRANTs defined. I'm a little puzzled at this point. Any ideas? Cheers, Arty -- 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
z/Linux image cloning
Hi, I'm also looking into cloning z/Linux images. I've been looking/using the available basic techniques, but I keep running into a problem. I suspect it has something to do with initrd. If my boot disk is R/O, I receive these messages prior to init running: Loading module dasd_fba_mod ... Using /lib/modules/2.4.21-231-default/kernel/drivers/s390/block/dasd_fba_mod.o dasd(fba): FBA discipline initializing dasd(fba): /dev/dasdb ( 94: 4),[EMAIL PROTECTED]: 9336/10(CU:6310/80) 1024MB at(512 B/b lk) dasd(fba): /dev/dasde ( 94: 16),[EMAIL PROTECTED]: 9336/10(CU:6310/80) 9MB at(512 B/blk) dasdb:CMS1/ LXSWAP(MDSK): dasdb1 dasde:CMS1/ LXSWAP(MDSK): dasde1 dasd(fba): We are interested in: Dev 9336/00 @ CU 6310/00 dasd(fba): We are interested in: Dev 3370/00 @ CU 3880/00 dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: EXAMINE 24: Command Reject dete cted - fatal error dasd(eckd): Sense data: dasd(eckd):device 0150 on irq 10: I/O status report: dasd(eckd):in req: 00ebef00 CS: 0x00 DS: 0x02 dasd(eckd):Failing CCW: 00ebeff0 dasd(eckd):Sense(hex) 0- 7: 80 02 00 00 00 00 00 00 dasd(eckd):Sense(hex) 8-15: 00 00 00 00 00 00 00 00 dasd(eckd):Sense(hex) 16-23: 00 00 00 00 00 a2 0f 00 dasd(eckd):Sense(hex) 24-31: 00 00 40 e0 00 00 00 00 dasd(eckd):24 Byte: 0 MSG 0, no MSGb to SYSOP dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: (EXAMINE) ERP chain report for req: 00ebef00 dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: 00ebef00: c5c3d2c4 dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: 00ebef10: 00eb ef00 00ebef00 dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: 00ebef20: 00ef 7800 00ec0e00 dasd_erp(3990): /dev/dasda ( 94: 0),[EMAIL PROTECTED]: 00ebef30: 00eb ... All the bind mounting and other magic happens *after* init runs. It looks like something is insisting that the boot disk be R/W during initrd. Any ideas? Cheers, Arty -- 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 under VM and Cloning
Hi, I've found William Sculley's basevol/guestvol approach to be fairly simple to set up. It's too bad that I goofed somewhere, as I can't get past my read-only root volume problems (with initrd). If I can get past this problem, the setup looks extremely promising (i.e., cloned images would only need 150cyl R/W minidisks as a starting minimum). Has anyone else run into problems with R/O root disks? Something in initrd wants that disk R/W. Adam suggested changing the parmline to something like "DASD=150(ro), ..." but once that is done, you lose the ability to update the basevol root disk (or a copy of it). Many thanks, Arty -- 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: LVM under Debian
Hi, I'd use: "vgdisplay -v lvmdata" and look for the "Total PE" value, and then use "lvcreate --extents nn -n vol1 lvmdata", where "nn" was the value from the "Total PE" display from vgdisplay. This way, I get exactly the size I want with no wasted space. Cheers, Arty -- 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
Swapping indicators
Hi, I'm working on a rush, special project that needs to determine if swapping has increased (as a result of doing something *interesting*). The program is written in Perl and so far, it looks at SwapFree and SwapTotal from /proc/meminfo. It computes a value of SwapUsed from these values and saves this value as a baseline. Later, the program does some *interesting* things and needs to determine if swapping has increased. So, it's off to /proc/meminfo again, recalculate SwapUsed, compare against baseline, and make an educated guess if any increased activity is the direct result of the program's Linux alterations. Yuch. Is there a "better way" to determine when swapping has increased? Are there other /proc files that might be useful? Trapping the output from vmstat for "si" and "so" values might be interesting, but certainly not clean (I'd rather find these values directly if I could). Cheers, Arty -- 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
vmhalt=LOGOFF
Hi, I know I've seen this before, but is it vmpoff=LOGOFF that causes a SLES8 guest (with /etc/inittab properly configured for "ca") to LOGOFF when it receives a SIGNAL SHUTDOWN? Or is it vmhalt=LOGOFF? I have both specified and something's not working. Thanks, Arty -- 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: Does TurboLinux do PAM ?
Hi, Does TurboLinux support LDAP via /etc/nsswitch.conf? I think Turbo is missing a package that makes all this work nicely with PAM. nss_ldap? Cheers, Arty
Re: RXSocket for REXX
On Thu, 7 Feb 2002 21:35:13 +0100 you said: >Todd Blevins wrote: >> >> Has anyone "ported" the rxsocket function (for rexx) to linux? > >Doesn't IBM's Object REXX for Linux have socket support ? Hi, Pierantonio Marchesini of CERN did a port, circa 1995 (!) to Regina. This URL should help: http://wwwinfo.cern.ch/regina/extention.html I lost touch with him, but his last contact information was: Pierantonio Marchesini C.E.R.N. ETH ZurichCH-1211 GENEVE 23 email: [EMAIL PROTECTED] Phone: +41.22.767.50.23 [EMAIL PROTECTED] Pierantonio made a few other nice additions to Regina. The source updates were in an AFS directory at CERN, but I never had a chance to request the address. IBM's OREXX has socket support, but the syntax is different than REXX/Sockets. However, there was a little presentation (SHARE?) describing how to do similar things using both packages. I have the presentation online, if anyone is interested. Cheers, Arty
R/O Linux guest?
Hi, I need to set up a Linux/390 guest with read-only file systems (and minidisks). Has anyone done anything like this? I'd want to put the portions of the file systems that MUST be read-write on a V-disk. My filesystems are all on separate minidisks: / /home /usr /opt /tmp /var swap I think I can make /usr, and /home read-only rather easily. The same with /opt. For the others, I'd have to disect them to locate the files that must be read-write and set up symlinks for them. Am I on the right track? Cheers, Arty
Re: Authentication on Linux using PAM to a z/OS RACF server
Hi, On Tue, 19 Mar 2002 10:19:07 EST Jim Elliott said: >> I didn't mean RACF, I meant DB2 under VM - there is no RACF involved > >Carlos: I was responding to Phil's question there. However, in looking >at Neale Ferguson's port of LDAP to VM (it was the old UMich code), it >does not appear to use any kind of external database. > >To get this to work on a z/VM base, someone would need to port OpenLDAP >to the z/VM Open Extensions ("UNIX") environment. Why port OpenLDAP to z/VM OE? Just leave it on a Linux Guest and instrument the guest to talk to RACFVM (via CP Diagnose x'A0' or a new API). Cheers, Arty
Re: send messages to vm console
Hi, You can use the logger command on Linux/390 and point the /etc/syslogd.conf at your VM host. You can then install my TCPLOGD server (on the 1997 VM Tools Tape) and customize it to your heart's content. The TCPLOGD "server" is a REXX EXEC and a NAMES file. That's it. Piece of cake. I've always thought that PROP should be taught how to listen for SYSLOG messages. In fact, I think it probably can be done without mods (just run something like TCPLOGD EXEC, with a few changes, in a background CMS process). Might be quite fun. If anyone has a ton of free time, I'll supply the tools. Cheers, Arty
Re: LINUX & Security
On Tue, 10 Dec 2002 15:06:39 -0500 David Boyes said: >> If you are an ACF2 (or CA-Top Secret) customer, then we have an >> open-source PAM plug-in that lets you authenticate directly >> against ACF2 >> or Top Secret. The client side (the part that runs on Linux) is >> available in source code or pre-built RPM form (both Intel >> and mainframe >> Linux). The server is simply a built-in integrated part of ACF2. With >> our plug-in installed, you need no user definition on Linux - your >> existing mainframe security rules and passwords are all >> that's needed. > >Nice. Will it be available for VM:Secure? How about RACF? Cheers, Arty
OT: New Mac G5
Hi, I'm considering replacing several existing development workstations with a single G5 Mac. The G5 would run Linux natively with the following applications: - Mac-On-Linux (for running Panther and Jaguar) - Hercules (to replace my aging P/390) - Bochs or Wine (for running those few Windows 2000 apps that have no native Linux client) - I would *LOVE* to be able to boot AIX on this thing, but I have no idea if that is possible Has anyone any experience with the G5? Any advice? Cheers, Arty P.S. I'd load the G5 to the gills with RAM and DASD.
Re: Use of DIrmaint
Hi, On Thu, 26 Feb 2004 17:29:57 -0500 David Boyes said: >Problem is that it's not helpful if you don't have a C compiler and a lot of >shops that are using VM only to support Linux aren't going to license the >C/C++ compiler just to write management apps. Can those folks running z/VM and Linux on IFL engines even license the IBM C/C++ compiler. Cheers, Arty
Using DCSS for extra memory
Hi, Many years ago, we created an application that would attach a R/W DCSS to use for extra memory. As the demands for more memory increased, subsequent DCSS space would be attached. We used to attach existing DCSS space (such as VTAM) for testing (no, this would no trash the real VTAM segment, we'd just ATTACH the DCSS in R/W mode, thus extending our virtual machine storage dynamically). I was wondering if Linux on 390 could be taught how to do this. Can Linux on 390 handle "holes" in it's address space? Cheers, Arty
Re: VSWITCH and LINUX
Hi, Two other collective nouns for penguins are "waddle" and "raft". A "waddle" is a bunch of penguins on land, and is preferred over "rookery". A "raft" is a group of penguins floating on the surface of water (where they spend most of their time). Cheers, Arty -- 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
Installing SLES8 31-bit on z990
Hi, I know I'm missing the obvious, but how do I force an install of a 31-bit SLES8 on z z990 (running unver z/VM, of course)? Cheers, Arty -- 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
gPXE for s390x?
Hi, I'm messing around with iscsi for a new project. I didn't have an x86 machine handy, so I ran some tests using VM guests. I found that RHEL 5.2 does not allow you to install to an iscsi target, and SLES10 SP1 installs just fine. Very nice. So, how do I boot from it? I assume I'd need to network boot gPXE or something that will set up an iscsi initiator to talk to my iscsi target. Any ideas? Is there a gPXE for s390x architecture? Our target systems for this project are a whack (tm) of x86 blade servers with no support for iscsi boot (just PXE->gPXE chaining magic). Cheers, Arty -- 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: Cobbler anyone?
Hi, I've been looking at getting cobbler to run on s390x to serve as a provisioning manager for x86 (and friends). It installs without too much fuss on RHEL 5.2. Getting it to install on SLES10 SP2 looks far more challenging. Our immediate need is for rapid x86 provisioning (we have the s390x provisioning down to a science). Cheers, Arty -- 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
LVM multipathing in SLES9
Hi, The "pvpath" command appears to have vanished in SLES9. We use "pvpath" and LVM multipathing for a dozen Linux guests running SLES8. Using SLES9, "pvcreate" works fine, but "vgcreate" complains about duplicate "PV" identifiers (on the alternate path devices). Although "vgcreate" indicates that the volume group was sucessfully created, it isn't. I have 8 paths to each FCP LUN on an IBM ESS. So, /dev/sd[b-h] are the same devices as /dev/sda. Works just fine under SLES8. I am able to define the LUN mapping in SLES9 using YaST2 LVM. I just can't convince "vgcreate" to do the right thing. Any ideas? Thanks, Arty -- 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: Bug fix in mksles9root
Hi, That SCSI underrun problem surfaced just after the 2.4 "-261" kernel level. I have several "-261" guests using SCSI and they do not have the problem. I have a guest at the "-281" level that has the problem, and I am certain that at least 1 earlier level (than "-281") also exhibited the problem. So, somewhere between "-261" and "-281", the problem was introduced. Recent 2.4 kernels ("-292" SP4) still have the problem. It has not been fixed. Cheers, Arty -- 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: Cloning FCP-attached SCSI SLES9
Hi, I've been using /usr/lib/YaST2/bin/YaST2.firstboot from a VNC session after cloning SLES9 images. I build a master SLES9 image, apply service and any generic packages, then I shutdown the image and clone. When I bring up a clone, it has the original hostname and IP address of the master. These get changed when I use firstboot from a VNC session. A subsequent reboot, and I'm ready to do another clone. I also use /usr/lib.YaST2/bin/YaST2.firstboot2 after I reboot the clone, but I haven't verified if it is really needed. Using firstboot lets me use a supported mechanism for changing all the values for a clone, without the guesswork. Cheers, Arty -- 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: 2005-10-04 Recommended Linux on zSeries code drop to developerWorks
Hi, On Wed, 5 Oct 2005 10:59:50 +0200 Christian Borntraeger said: >They differ in terms >of response buffer handling(vmcp tells you if the buffer was too >small by returning an error code instead of increasing the buffer >and retry) That's really too bad. Would an end-user have control over the size of the buffer? I hope not (that would be severely bad form). Automatically increasing the buffer size and retrying is "The Right Thing To Do". Why not adopt cpint and be done with it? Why re-invent the wheel? Cheers, Arty -- 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: DIAG [was: 2005-10-04 Recommended Linux on zSeries ...]
Hi, If good IUCV support was available, *SPL would be rather cool. It's better than DIAG14 because it is a defined asynchronous interface. DIAG D8 might also be a fun addition to Linux. Cheers, Arty -- 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
FCP LUNs on SLES9 SP3+
Hi, I have a need to add a great many FCP LUNs to a SLES9 guest. Under SLES8, it was a simple /etc/zfcp.conf file. With SLES9, is it more than building a bunch of /etc/sysconfig/hardware/hwcfg-zfcp-bus-ccw-0.0. files? I'd really, really, really hate to go through the YaST2 panels to do this. Especially since I need to define 4 (or was it 8?) paths to each LUN. I'd like to adapt my existing zfcp.conf REXX EXEC to generate the hwcfg files instead. Cheers, Arty -- 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
FCP LUNs on SLES9 SP3+
Hi, So it turns out, that if you use the YaST panels to add the zfcp driver, then it is a simple matter of dropping a bunch of hwcfg configuration files in /etc/sysconfig/hardware. A simple REXX EXEC makes light work of generating the hwcfg files (with 8 paths to each LUN!) After a reboot, the LUNs are found. The next step to save myself a lot of grief is to decipher the EVMS CLI syntax and create a shell script to batch-define the EVMS objects. Cheers, Arty -- 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: Implementing zFCP & Multipathing on existing SLES9 SP3 with ECKD/LVM ...
On Tue, 6 Jun 2006 12:33:35 -0400 Terry Spaulding said: >Has anyone been able to have a coexistence with ECKD LVM non Multipathing >and SCSi with LVM Multipathing ? We use ECKD (non-LVM) and SCSI (LVM Multipathing) by changing /etc/evms.conf as follows: - Under "legacy_devices", remove "dasd?" from "include = ..." and add "dasd*" to "exclude = ..." - Under "sysfs_devices", add "dasd*" to "exclude = ..." I would expect that you would only need to do our second change to evms.conf in order to use ECKD (LVM non-multipathing) and SCSI (LVM Multipathing). But that's just a guess. We only use LVM with SCSI. And we always multipath SCSI. Cheers, Arty -- 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
NED
Hi, Has anyone thought of adding a DIAG58 driver to ned? It would remove the TERMINAL CONMODE 3270 requirement. Cheers, Arty -- 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
/dev/random
Hi, I tried to use gpg today to generate a pgp key on SLES9x. The process hangs while reading from /dev/random (which is lacking entropy). Hasn't this whole /dev/random (and /dev/urandom) thing been beaten to death? Does anyone else use gpg on s390x? Cheers, Arty -- 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: /dev/random
Hi, On Mon, 21 Aug 2006 18:58:38 -0400 Post, Mark K said: >Not specifically, but I just checked one of my SLES9 SP3 s390x systems: ># cat /proc/sys/kernel/random/entropy_avail >2944 > >My system is fairly idle. What are you running on your system?=20 My guest is very very idle. The z990 is somewhat large. entropy_avail = 355 Cheers, Arty -- 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: Best Swap Disk Strategy?
Hi, On Fri, 29 Sep 2006 16:35:30 -0500 Mark Wheeler said: > >I have 5.2 GB of vdisk allocated to 30 servers. We have about 50G of vdisk allocated to 50 or so servers. I've completely removed the cap on vdisk allocation. We've also had great success using cmm to "hide" much of a server's memory, and dynamically adjusting the memory size based on swapping pressure. Yeah, we could DEFINE STORAGE and re-IPL, but why take an outage? Then again, z/VM 5.2 has completely spoiled us and we no longer fear very large virtual machine sizes. Cheers, Arty -- 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
Installing SLES9 to multipathed FCP SCSI
Hi, I've been trying to install SLES9 to a multipathed FCP SCSI LUN. I can figure out how to get the 4 paths defined, and I can get SLES9 installed without multipathing. I normally use evms for FCP SCSI path multipathing, but I can't figure out how to get evms to add the requisite regions during the install process. I'd hate to have to pre-build the evms segments prior to the install process. Anyone have any clues? Cheers, Arty -- 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
getgroups for root
Hi, I'm running into a strange problem with SLES9. We are doing a TIM/TAM rollout and a software component (MQ) is complaining during startup. It seems that the user that runs the start script (root, during boot in our case) does not appear to be in the mqm group. An "id" command issued from the 3215 console confirms. A "cat /etc/groups" contradicts, as does a "sudo id". If we run the start script from an ssh session, it works (as does the "id" command). It turns out that "login" might be the culprit, as it only sets the primary group if the uid is 0. I tried removing minigetty from /etc/inittab, and tried "sulogin", Rick's suloginv script, and /bin/sh all to no avail. I really don't mind login's behaviour, but the rc scripts run before login, so shouldn't root's list of auxiliary groups be set by getgroups properly during boot? Cheers, Arty -- 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: getgroups for root
(whoops, replied off-list) Hi John, mingetty on the 3215 console runs through "login". $HOME is set, as other environment variablesb, but "id" still behaves oddly. For the rc script case, I would not assume any settings from "login", I did assume that "getgroups" had been issued for root. Cheers, Arty -- 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: AF_IUCV support
Hi, On Wed, 28 Mar 2007 22:32:23 -0400 David Boyes said: >> So how about if you make the calls to the ACI stubs nice CP exits so >> that we can dynamically load the modules that RACF (or someone else) >> supplies. > For what its worth, our home-brew SECURE system is implemented as a CPXLOADable IUCV System Service. The dynamic IUCV System Service code is a local CP modification, but wouldn't be needed if *RPI spoke to CP exits. In any event, CPXLOADable security code works. Well. Cheers, Arty -- 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
SLES9 and EVMS
Hi, We use EVMS on SLES9X to handle our scsi multipathing needs. We pretty much use the procedure documented in SG24-6344 Linux for zSeries: Fibre Channel Protocol Implementation Guide, page 192-234. We have 4 paths defined for each LUN. We had a failure on our SAN Volume Controller this weekend that caused EVMS to lose 2 disks (child objects). These 2 disks represent 2 of 4 paths to the same LUN. EVMS complains that the disk count in the "region" is wrong (ok, we lost 2 disks), but it won't continue without the missing disks (even though these disks were defined in EVMS as being redundant paths to the same LUN). It seems odd that EVMS would work this way. Would anyone know how to "fool" EVMS into working with missing child objects? Shouldn't EVMS handle path failures? Novell tells me that multipathing with EVMS is not a supported configuration. I thought it was supported for non-root volumes. Sigh, Arty (35 dead penguins) -- 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