Re: RACF Console Logs
On Oct 26, 2009, at 4:32 PM, Schuh, Richard wrote: That's the way IBM delivers it :-) In the case of OPERATOR, watch out for OPERATIONS authority - some people naiively give OPERATOR too much authority. Do you leave its password set to OPERATOR too? I mean, after all, that's the way IBM delivers it Adam
Well, *finally*!
On my very very last day at work at Sine Nomine, guess what I finally got, courtesy of David and Margarete? That's right. A pony. Actually, quite a few ponies. http://www.flickr.com/photos/17339...@n00/3996384848/ I'll attach it too, but I think that will get stripped--so use the flickr link. Adam P.S. So *there*, Chucky.
Re: z/VM Linux on Cp
On Oct 7, 2009, at 9:18 AM, Wakser, David wrote: Adam: Please explain, for those of us not yet involved in Linux, why it's not cost effective. For example, if we already have z/VM running, there is no additional cost involved. David Wakser Maybe I'm undercaffeinated. Since it's in its own LPAR, you would have to pay standard-engine licensing fees, but if Linux is the only thing in that LPAR, then it doesn't matter. So never mind. I was thinking that CP-versus-IFL would mean that it would drive up your other software costs but as its own LPAR, I guess not. The cost issue--and the reason to run specialty engines--is simply that you do not want to pay standard engine licenses for engines that are not running traditional IBM mainframe workloads. Adam
Adam is moving on...
Some of you know this already, and some of you don't: Last month, I was offered a great opportunity at another organization that will allow me to focus my career in a way I have been interested in for quite a while. I accepted that position, knowing that my customers at SNA would be in good hands after I left. As a result, this Friday is my last day with SNA. Sine Nomine and its owners have been quite supportive of my move, for which I am very grateful. I intend to continue reading the Linux on 390 and VM mailing lists, but I am no longer going to be working with zSeries boxes as part of my job (I will continue to play with Hercules in my Copious Free Time), so my advice may grow even less useful than it currently is. Anyone who wants a non-work email address of mine and doesn't already have one, please contact me off-list. Adam
Re: LOGOFF/FORCE PENDING
On Oct 1, 2009, at 2:46 PM, Schuh, Richard wrote: After several hours, but before a dump could be taken, Tried metamucil? Adam
Re: VM lockup due to storage typo
On Sep 18, 2009, at 9:11 AM, David Boyes wrote: I think we're all in violent agreement on that point. Now, the question is what is the best way to put a safety on that gun? Oooh! Oooh! Pick me! Mandatory User Access Control dialog boxes that pop up and make you click OK any time you want to breathe. Adam
Re: VM lockup due to storage typo
On Sep 17, 2009, at 1:58 PM, Bill Holder wrote: I'd agree with that point in cases where it's less clear, but in this case, it's perfectly clear that the user action would have been harmless if not for the administrator typo Yabbut Administrator typo is not a failure mode the operating system is designed to protect you from. If you have authority to edit the user directory, then, well, your gun, your foot. Adam
Re: VM lockup due to storage typo
On Sep 17, 2009, at 5:36 PM, David Boyes wrote: Whether it should march off a cliff without at least questioning the order is the question at hand. Of course it should. Yes, my Unix is showing. Adam
Re: Download Vm Tape
On Sep 2, 2009, at 9:37 AM, Alan Altmark wrote: I don't recommend going into production with any system that has default dasd labels or passwords. But that's just me. Passwords I can see. But changing DASD labels? How come? If it's security-through- obscurity, then it's not very obscure, since it's pretty easy to figure out where the res volume actually isso there must be some other reason. What is it? Adam
Re: Retrieving a VM Packed file
On Sep 2, 2009, at 1:35 PM, David Boyes wrote: I think my list of “must-haves” would be: VMARC the PIPE-friendly DDR TRACK plastic pipes XCOL CHARLOTT *I* really like CUA2001, but, uh, that's just my perversity. Adam
Re: Download Vm Tape
On Sep 2, 2009, at 3:13 PM, Schuh, Richard wrote: I think it goes more to separating production from installation/ maintenance. It the volsers are different, changing something on the system must be deliberate. Accidental update is very difficult. Alan is big on data integrity, well ... as big as he can be. Butunless you're installing your second-level systems on dedicated devices, would you ever have duplicate volsers? And, uh, why would you install second-level like that unless you really, really meant it? Adam
Re: How to tell how many linux running on z/VM?
On Aug 14, 2009, at 11:42 AM, Alan Altmark wrote: IPL 190 PARM NOSPROF INSTSEG NO Who am I? :-) With that line, most likely Chucky. Adam
Re: z/VM 6.1 and Hercules on Z9
On Jul 29, 2009, at 8:48 AM, Edward M Martin wrote: Hello Adam, I can say that I was told, in no uncertain terms, from IBM that it is illegal to run VM/ESA and up under Hercules. That was a while back, and things may have changed but I have never seen anything to the contrary. Hey, IBM what is your stand on Hercules? Your answer was probably complicated by the fact that when you asked the question, IBM very likely assumed an implied on commodity hardware. So I think we need to ask a followup question: ...on Hercules, when running in a z/Linux image, underneath an instance of z/VM on the processor or processors to which that version z/VM is licensed. Certainly no one ever told me what I was doing on my H70 was illegal. A lot of people asked me questions that implied that they thought I was an idiot or at least insane--which, you know, from a performance standpoint, well, yeah, guilty as charged--but no one said that I shouldn't run my copy of z/VM at about 1/100th speed if that was what I really wanted to do. Adam
Re: z/VM 6.1 and Hercules on Z9
On Jul 29, 2009, at 1:50 AM, Dave Wade wrote: I seem to recall it being mentioned somewhere that after the PSI debacle IBM amended the license terms to specifically prohibit the use of zVM under Emulation of any kind thus closing this loophole. Even if it's on the same processor? That would surprise me. After all, you are already allowed to nest your z/VM instances as deep as you like (if this were to change, it would break pretty much everyone's upgrade process, wouldn't it?). Whether or not some of the intervening layers are Linux doesn't seem like it should matter, but strange are the ways of licenses. Adam
Re: z/VM 6.1 and Hercules on Z9
On Jul 28, 2009, at 4:25 PM, Edward M Martin wrote: I can say that, 1) it is illegal to run z/VM (pick a version) under Hercules, 2) they do not like it, and 3) they do not have any sense of humor. Perfectly fine to run it under Hercules on the processor z/VM is licensed to. Adam
Re: z/VM 6.1 and Hercules on Z9
On Jul 28, 2009, at 6:32 PM, John McKown wrote: On Tue, 28 Jul 2009, Adam Thornton wrote: On Jul 28, 2009, at 4:25 PM, Edward M Martin wrote: I can say that, 1) it is illegal to run z/VM (pick a version) under Hercules, 2) they do not like it, and 3) they do not have any sense of humor. Perfectly fine to run it under Hercules on the processor z/VM is licensed to. And how would you get IBM to license z/VM 6.1 to you when they know that you don't have a z10? IBM is not under any obligation to license their software to anybody. Now, if IBM ever said NO simply because we don't like you, that might be a very interesting lawsuit. I'm not denying that the original poster may not have a legal way to do it. I am specifically objecting to Edward Martin's Point 1. There's a perfectly legal way to run z/VM under Hercules. Not that you'd necessarily want to. But I did run 64-bit z/VM 4.4 on my H70, and I might be the only person who can say that. Adam
Re: REXX and Panels
On Jul 13, 2009, at 6:46 AM, Thomas Kern wrote: Many years ago, IBM had a package, I think it was called REXX/CUA. I am not in the office today or I could give you the product number. We wrote several menued execs using it. I don't remember the price, if there was a price. It was ALL in REXX and XEDIT. Good news: you're describing CUA2001: http://www.vm.ibm.com/download/packages/cua2001.vmarc On the VM Downloads Page now. I wrote a lot of stuff in it at Rice when it was still a paid product; I've messed with it a bit more since then. It's a pretty easy way to do an XEDIT interface for a Rexx ap. Adam
Re: PERFSVM question
On Jul 8, 2009, at 11:15 AM, David Boyes wrote: Simple answer: put a Linux guest in front of the VM TCP stack with the old address as the external address, renumber the VM stack to a RFC1918 address on an internal guest lan, and enable IP Masquerade in iptables. That gets you all sorts of useful info, and lets you shut them down cold. Add one of the IDS toolkits, and you can clobber the twerps network wide. -Original Message- From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf Of Jim Bohnsack Sent: Wednesday, July 08, 2009 11:02 AM To: IBMVM@LISTSERV.UARK.EDU Subject: PERFSVM question We saw a bunch of logon attempts a night ago to userid ADMINIST which I do not have defined in the directory. There were about 2,500 over the course of 2 hours. They were apparently not coming in thru an emulator, so that pretty much leaves the web interface to Performance Toolkit. Is there any way I control that interface. How can I get the ip address? IBM used to have, internally, a mod that would double the amount of time between each unsuccessful logon attempt to a particular userid. Something like that would do the job. Are you running an FTP server? I saw an attack on a system using that userid (well, Administrator) coming in via FTP a few weeks ago. Adam
Re: PERFSVM question
On Jul 8, 2009, at 11:15 AM, David Boyes wrote: Simple answer: put a Linux guest in front of the VM TCP stack with the old address as the external address, renumber the VM stack to a RFC1918 address on an internal guest lan, and enable IP Masquerade in iptables. That gets you all sorts of useful info, and lets you shut them down cold. Add one of the IDS toolkits, and you can clobber the twerps network wide. -Original Message- From: The IBM z/VM Operating System [mailto:ib...@listserv.uark.edu] On Behalf Of Jim Bohnsack Sent: Wednesday, July 08, 2009 11:02 AM To: IBMVM@LISTSERV.UARK.EDU Subject: PERFSVM question We saw a bunch of logon attempts a night ago to userid ADMINIST which I do not have defined in the directory. There were about 2,500 over the course of 2 hours. They were apparently not coming in thru an emulator, so that pretty much leaves the web interface to Performance Toolkit. Is there any way I control that interface. How can I get the ip address? IBM used to have, internally, a mod that would double the amount of time between each unsuccessful logon attempt to a particular userid. Something like that would do the job. My FTP attacks were coming from somewhere inside Korea Telecom, if that's of any use to you. Adam
Re: PERFSVM question
On Jul 8, 2009, at 12:03 PM, Jim Bohnsack wrote: Easy for you to say. How about She sells sea shells by the seashore? Adam
Re: CP Query wildcards
On Jul 1, 2009, at 2:33 PM, Richard Troth wrote: By the way ... Unix cheats. The shell expands all wildcards, which I have always said is a mistake because it presumes on the context. The shell can only expand wildcards that are filenames. Not everything you might want to wildcard is a file. Wouldn't it be nice if you could 'ifconfig eth*'? Wouldn't it be nice if Unix worked like Unix was supposed to work, and everything *WERE* a file? Yeah, yeah, I know, Plan 9 is right over there if I want it. Adam
Re: Last release for 3420s?
On Jun 5, 2009, at 1:08 PM, Stephen Frazier wrote: Some people who have tried 3420's on z/VM 5.3 said they work. However IBM is no longer testing 3420 code and if it ever stops working they may not fix it. Finding the actual 3420 hardware that still works is getting difficult. :) There's some in Virginia free to a good home. Free to any home, in fact. Just take it away. As long as you provide all the labor and all the transport, it's yours. Not kidding. Adam
Re: HCPDDR704E error attempting to copy res volume
On Jun 4, 2009, at 9:21 AM, Edward M Martin wrote: Typically Maint has 540RES as 123 MR. I should not answer questions before coffee. Yeah, there already *IS* a covering minidisk, isn't there? Question: is it actually safe to DDR the RES volume from a live system to another system? I always *thought* that was one of those things that you usually got away with but weren't supposed to do, but I will admit to perhaps being conditioned by growing up in the Unix world, where doing that kinda stuff with mounted filesystems is a Bad Idea. It would be very nice to know if it is in fact not risky; it will save me the time of going back after I've done all the user volumes, shutting down the system, and using standalone DDR on the RES volumes. Adam
Re: HCPDDR704E error attempting to copy res volume
On Jun 4, 2009, at 9:14 AM, Bauer, Bobby (NIH/CIT) [E] wrote: I’m attempting to clone our pilot VM system to a production lpar since it looks like a couple of applications might actually have a real future (a quiet hurray!). Attempting to copy the res volumes fails q 127a DASD 127A CP OWNED 540RES 83 then ddr input 127A 3390 540RES HCPDDR704E DEVICE 127A NOT OPERATIONAL Any suggestions would be helpful. #1, define a covering minidisk from a user with DEVMAINT privs. #2, have the primary system down. Your RES volumes probably shouldn't be DDRed while in use. As with Linux guests and an unclean shutdown you will probably get away with with after a FORCE start when you come up in production, but, well, you're going to production! Since #2 makes #1 hard, maybe a better solution would be to IPL standalone DDR and do it that way. Adam
Re: Clean Linux Guest Shutdown
On Jun 3, 2009, at 12:43 PM, Robert J McCarthy wrote: I am trying to develop a shutdown procedure to cleanly shutdown my linux guests, prior to shutting down vm. Reading the documentation in the virtualization cookbook for SLES10 and the vm CP COMMANDS manual; I have setup the following : 1. In each linux guest's /etc/inittab; I have changed the shutdown - r to shutdown -h 2. In my autolog1 exec I have placed the following command: CP SET SIGNAL SHUTDOWN 1200 ( To allow the guests 20 minutes to respond) Note: I have also entered the command manually When I issue the shutdown, vm shuts down before most if not all linux guests have responded or completed shutdown; always within a minute or two. As a result I end up with file corruption in some linux guests after vm is re-IPLed and the guests are brought back up. Is there a better way to accomplish a clean linux shutdown. Thank you, Bob Our SYSVINIT drop-in-replacement for a list-of-machines-in-autolog would do the trick. It may be overkill. Adam
Re: What we must do before we claim the zlinux server is in production stage?
On Jun 2, 2009, at 11:51 AM, Tom Duerbusch wrote: A lot of it also depends on local practices. 1. Backupsscheduled..and monitored. And RESTORED, whether you need to or not, on some schedule. A good test, I'd say, is to pick ten files at random from the backup catalogue every so often and restore them to a temporary location, and then verify those files. (Assuming you can spare your tape library long enough, because ten random files is a lot of loading/unloading/ seeking.) Seriously: your backup regimen is USELESS if you cannot restore the files you backed up, and when you need them is NOT the time to find out that the tapes haven't been being written correctly. Adam
Re: What we must do before we claim the zlinux server is in production stage?
On Jun 2, 2009, at 7:57 PM, Martin, Terry R. (CMS/CTR) (CTR) wrote: Yes, just having some fun! It's not Friday, so fun is not permitted here. Move along. Adam
Re: Any idea? Dirmaint error by detach 123 disk(540RES)
On May 29, 2009, at 6:31 PM, Alan Altmark wrote: I really must get cracking on the z/VM Auditor's Field and Survival Guide [Free cudgel included at no extra charge! Now contains the most common phrases heard from z/VM security weasels in the wild, including such hits as I'm gonna whap you upside yo' head, son! and I don't think so, Tim. For a limited time only, it also includes a fold-out full- color diagram of the interior of the human kneecap.] Put me down for a dozen. Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 2:14 PM, Miguel Delapaz wrote: Adam, What does NETSTAT DEV say? Well, *IT* says that my interface there is not working, and TCPIP startup bears this out: Device ETH1Type: OSDStatus: Inactive Queue size: 0 CPU: 0 Address: 7008Port name: UNASSIGNED IPv4 Router Type: PrimaryArp Query Support: No IPv6 Router Type: Primary Link ETH1 Type: QDIOETHERNET Port number: 0 Transport Type: IP Speed: 1 BytesIn: 0 BytesOut: 0 Forwarding: Enabled MTU: 1500IPv6: Configured IPv4 Path MTU Discovery: Disabled Broadcast Capability: Unknown, but previously not available Multicast Capability: Unknown The relevant bit of TCPIP startup is: 14:55:58 DTCOSD066E OSD device ETH1: ToOsd ProcessReadBuffer: Termination failur e code F6 14:55:58 DTCOSD355E OSD device ETH1: Possible LAN transport misconfiguration det ected during OSD device initialization. 14:55:58 DTCOSD082E OSD shutting down: 14:55:58 DTCPRI385IDevice ETH1: 14:55:58 DTCPRI386I Type: OSD, Status: Inoperative 14:55:58 DTCPRI387I Envelope queue size: 0 14:55:58 DTCPRI497I Address: 7008 Port Number: 0 HELP DTCOSD066E doesn't seem to help. Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 2:29 PM, Miguel Delapaz wrote: Adam, Transport Type: IP This says your NIC is defined as Layer 3. 14:55:58 DTCOSD355E OSD device ETH1: Possible LAN transport misconfiguration detected during OSD device initialization. This says you tried to attach a Layer 3 NIC to a Layer 2 LAN (or vice-versa) Indeed. I defined it to TCPIP with DIRM NICDEF and I don't see any way to specify whether I mean Layer 2 or Layer 3. I just defined it as QDIO. I think my cough syrup is failing me. How do I tell DIRMAINT, no, really, QDIO *ETHERNET* ? Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 3:22 PM, O'Brien, Dennis L wrote: Adam, You don’t specify layer 2 or 3 on the NICDEF. You specify it on your DEFINE LAN or DEFINE VSWITCH statement. Yeah, but I DID that: Here's the L3 LAN: LAN SYSTEM GLAN1Type: QDIOConnected: 1Maxconn: INFINITE PERSISTENT UNRESTRICTED IPAccounting: OFF Here's the L2 LAN: LAN SYSTEM L2LANType: QDIOConnected: 2Maxconn: INFINITE PERSISTENT UNRESTRICTED ETHERNET Accounting: OFF I'm *pretty* sure (but not positive) that the lan definitions predated adding the NICs to TCPIP Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 3:24 PM, Miguel Delapaz wrote: I defined it to TCPIP with DIRM NICDEF and I don't see any way to specify whether I mean Layer 2 or Layer 3. I just defined it as QDIO. I think my cough syrup is failing me. How do I tell DIRMAINT, no, really, QDIO *ETHERNET* ? Huh...that appears to be an interesting oversight on our part. I've forwarded your note off to the dirmaint developer for verification, but it doesn't look like it's possible at this point. Um, OKthen how do I specify it in the directory entry? I don't mind doing a DIRM GET, editing the damn thing, and then doing a DIRM REPLACE.but what do I edit it *to* ? Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 3:36 PM, David Kreuter wrote: Why are you picking on poor ole DIRMAINT? It's CP that doesn't support it in the NICDEF statement! not DIRMAINT - OK, so, look, surely I'm not the first person to want to do this. How do I couple a virtual machine's virtual NIC to a Layer 2 Guest LAN? How do I define a Virtual NIC for a virtual machine that is a Layer 2 device? CP Q V 7000 OSA 7000 ON NIC 7000 UNIT 000 SUBCHANNEL = 7000 DEVTYPE OSA CHPID 00 OSD 7000 MAC 02-00-00-00-00-01 CURRENT 7000 QDIO-ELIGIBLE QIOASSIST NOT AVAILABLE CP Q V 7008 OSA 7008 ON NIC 7008 UNIT 000 SUBCHANNEL = 0006 7008 DEVTYPE OSA CHPID 02 OSD 7008 MAC 02-00-00-00-00-03 CURRENT 7008 QDIO-ELIGIBLE QIOASSIST NOT AVAILABLE They look pretty much the same. Same error here (and I did just try deleting the NIC and then adding it back in case I needed to have the L2 LAN precreated...): 16:41:30 DTCOSD066E OSD device ETH1: ToOsd ProcessReadBuffer: Termination failur e code F6 16:41:30 DTCOSD355E OSD device ETH1: Possible LAN transport misconfiguration det ected during OSD device initialization. 16:41:30 DTCOSD082E OSD shutting down: 16:41:30 DTCPRI385IDevice ETH1: 16:41:30 DTCPRI386I Type: OSD, Status: Inoperative 16:41:30 DTCPRI387I Envelope queue size: 0 16:41:30 DTCPRI497I Address: 7008 Port Number: 0 Adam
Re: Was I confused? L2 Guest LAN, z/VM 5.4
On May 15, 2009, at 4:09 PM, Miguel Delapaz wrote: OK, so, look, surely I'm not the first person to want to do this. With the ETHERNET option on the QDIOETHERNET LINK statement (in case you missed my other note) Yay! Thank you! (Your other note hadn't arrived quite yet.) I now have networking working to my Layer 2 guest. I am thrilled. Adam
Re: LTO4 tapes attached via FCP
On May 14, 2009, at 3:27 PM, Marcy Cortes wrote: I wonder if anyone is attaching a LTO4 (Linear Tape Open, Generation 4), via FCP, to a z9 or z10 running zVM 5.4? Yes, to SLES 10 SP 2 on both a z9 and z10 under zVM 5.4, via NPIV to I10K SAN switch. And if yes, are there any zVM utilities (DDR, SPXTAPE, etc) that support LTO4s? Don't think so. Or does zLINUX even support LTO4s? Certainly does. We're using them in an IBM ATL (3584 I believe is the type). With a bit of development effort, or some penalty of network/memory/ CPU overhead (i.e. going the NFS route--which could use an internal Guest LAN) you could certainly do CMS filesystem backups through-- well, *my* choice would be Bacula--a Linux guest to LTO tape. Neale put together a minimalist Bacula client for CMS back in the Bacula 1.3x days; with some work that could be ported forwards. I think as he did it it only knew about minidisks rather than SFS though. Adam
Re: Shared File System Interface
On May 3, 2009, at 3:39 AM, Malcolm Beattie wrote: David Boyes writes: On 5/1/09 4:20 PM, Alan Altmark alan_altm...@us.ibm.com wrote: Morituri te Salutant! :-) Dulce et decorum est pro patria mori, sed dulcius pro patria vivere, et dulcissimum pro patria bibere. Ergo, bibamus pro salute patriae. Bibit hera, bibit herus, bibit miles, bibit clerus, bibit ille, bibet illa, bibit servus cum ancilla... Bibo, ergo sum. Adam
Re: SWAPGEN
On Apr 26, 2009, at 9:44 PM, Martin, Terry R. (LOCKHEED MARTIN Performance Engineering/CTR) (CTR) wrote: Hi I am using SWAPGEN to define by z/Linux VDISKS I also want to define a real disk for swap. My question is can I use SWAPGEN to define a swap on real DASD? If you have an example of the control card syntax to accomplish this that would be great? Why bother? Since real DASD is persistent, once you make a swap disk in Linux the first time, you just keep it around. The reason for SWAPGEN is that VDISK goes away on logout, and it was a pain to modify the boot process to add the swap signature before a swapon -a is done. Adam
Re: Secure FTP
On Apr 4, 2009, at 11:19 AM, Alan Altmark wrote: On Saturday, 04/04/2009 at 12:10 EDT, Chip Davis c...@aresti.com wrote: On 4/3/09 17:29 Alan Altmark said: It was a tupo. Wow, that's impressive! tupo by itself is a 'meta-typo' but coupled with its reference to two releases of z/VM, that makes it a 'meta-typo pun'. Definitely Friday-level work, there! :-)) It just shows you what decades of study, contemplation, and bear claws** can achieve. Kind of brings a tear to your eye, doesn't it? (sniff) Alan ** a type of pastry with no redeeming social value whose only purpose is to carry fat, sugar, cinnamon, and a few (sugar-soaked) raisins into your body, where it has a half-life of 4 years. M'mmm m' good! Oh. I thought it was the pain brought on by the heart attack (in turn brought on by the clogged arteries from the years-of-bear-claws) that was bringing the tear to my eye. Adam
Re: Problem with PEEK Command - update
On Apr 3, 2009, at 1:50 PM, Raymond Noal wrote: I would like to thank all of you who responded. I was really surprised at how the majority of you made the leap from using PEEK to TCPIP data buffers. Well, it's highly intuitive, after all. Adam
Re: Guest Billing
On Apr 2, 2009, at 9:15 AM, =?iso-8859-1?Q?Greg_Dyrda?= wrote: We currently bill for Linux on a per guest basis. I'm wondering what approach others are taking. Specifically, I'm wondering if it is possibl e to bill at the process level and if anyone else is billing that way. If you want to go this way, you certainly want to bill from the VM processor utilization standpoint. Which won't actually know about the Linux process lists, but WILL tell you about actual consumption of physical resources, which is presumably what you really want to do anyway. Adam
h3270 Phun Phact
If you have h3270 pointing at a recent s3270, then you can get SSL support for free by specifying: L:hostname:portnum instead of just hostname in the connect to field. Other Phun Phact: you're probably going to have to edit your Tomcat (or whatever) policy to allow the h3270 program to execute the s3270 executable. Once you've done that, though, it's really quite straightforward. All h3270 is is a little Java web app built on top of s3270 as a screen- scraper. It works rather nicely. I would imagine that with a little clever css you could even use proper 3270 fonts and colors, although I haven't actually bothered yet. Also, wrap your Tomcat in SSL (gee, that sounds dirty!) so that you're not exposing (gee, that sounds dirty!) your password in the web part of the session (even if you have SSL to the host, unless you have Tomcat protected by SSL you're still sending username/pw in the clear to the web interface). Adam
Re: New CMS based SSLSERV problem... DTCSSL300E
On Mar 19, 2009, at 8:57 AM, Alan Altmark wrote: On Thursday, 03/19/2009 at 08:39 EDT, Mrohs, Ray ray.mr...@usdoj.gov wrote: Thanks Alan. Unfortunately our site is standardized on the Rumba client, and the centrally managed upgrades happen once a blue moon. It looks like it might be a while before we can utilize the new SSLSERV, even under the best circumstances. Is there a list of clients that have been tested and work? - IBM Personal Communications 5.9 works - Seagull's BlueZone works - x3270 works - wc3270 (Windows version of x3270) works - Zephyr Passport works - IBM Host on Demand fails - Micro Focus Rumba fails - Attachmate Reflection fails Has anyone tried tn3270X ? I'll be giving it a shot once I have the 5.4 SSL support working, but other stuff is likely to conspire to slow that down. Adam
Re: Please tell me I did something stupid
On Mar 18, 2009, at 3:10 AM, Rob van der Heij wrote: Bad idea to throw them away. But I would be happy with a summary chapter in the book that has the scenario for those who are not completely new to z/VM installation. It strikes me that, given that we have several versions' worth of excellent models, there's nothing really stopping, um, one of *us* from doing a set of one page versions, one for tape, for first level DVD, and for second level DVD. Of course, getting those KNOWN to people who don't already know about the mailing list becomes the trick. Adam
Re: Please tell me I did something stupid
On Mar 18, 2009, at 10:13 AM, Alan Altmark wrote: Checklists and other preparatory advice, based on real-world experience, are ponies of a different color. They would definitely be a value- add, providing *guidance* (opinion) where IBM can usually only provide a *procedure*. Hey, man, *you're* the one threatening to take away my pony. I don't care if the Quick And Dirty Guide is four or five pages, or whether there's a paper version at all. MY use case for the thing is to open it in one window, have my 3270 emulator in another window, and look at the guide when it's time to do the next step. Adam
Re: z/VM 5.4 Installation from DVD Failure - PJBR
On Mar 17, 2009, at 5:56 AM, Rob van der Heij wrote: Are you sure the disk is R/W and that it is the right size? I tried to do it from the quick ref guide and found that it was missing some of the steps that you need (and are only in the real book). You mean the missing 22cc and 2cf1 disks, or something else? That *was* sort of annoying, wasn't it? Not hard if you already have an idea of what is being done to construct the 2d-level primer system, but if you were a newbie, I can see how it would leave you baffled. And of course if you are (shock, horror) editing your user directory by hand then just adding the appropriate disks could be much harder than it was for me. If it hasn't been done already, could MAINT's default directory entry please be updated to include , 2CF1, 22CC ? Surely the HOPE is that a z/VM customer will go on to install a later version someday Adam
Re: z/VM 5.4 Installation from DVD Failure - PJBR
On Mar 17, 2009, at 11:23 AM, Rob van der Heij wrote: It really does not have to be in MAINT - you don't want to run the 2nd level system in MAINT anyway, so why mess with things. It's also only needed briefly in the process, so for me 3 T-disks would be fine. And in theory when you ship the GA of z/VM 5.4 you don't know upfront how large the payload for those disks will be in the next release... And I don't think you really need the three disks - it could have been enough just to ship a single disk to bootstrap the process. I would be surprised if /22CC ever needed to be anything other than 5 cyls. I made a 120-cyl 2CF1 because I assumed it was a shadow CF1 and that's 120 cyls on my 5.3 system. I agree that T-disk would be fine, and, yes, it seems as if putting that all on the same disk would make life easier. Adam
Please tell me I did something stupid
I just installed a virgin z/VM 5.4 with all products on filepool rather than on minidisk. I only tweaked it enough to get a TCPIP stack up and running so that I could FTP the RSU over to it, DETERSE it, and apply service. I've done literally nothing else to the system. 0 * * * Top of File * * * 1 2 SERVICE USERID: MAINT 3 4 Date: 03/17/09Time: 15:30:04 5 6 ST:VMFSRV2195I SERVICE ALL RPTF0168 7 ST:VMFSRV2760I SERVICE processing started 8 ST:VMFSUI2760I VMFSUFIN processing started 9 ST:VMFSUI2760I VMFSUFIN processing started for product 5VMLEN40%LE 00010 ST:VMFSET2760I VMFSETUP processing started for SERVP2P LE 00011 ST:VMFSET2204I Linking MAINT 4C4 as 4C4 with the link mode MR 00012 SV:HCPLNM107E MAINT 04C4 not linked; not in CP directory 00013 SV:VMFSET1965E The command, CP LINK MAINT 4C4 4C4 MR , failed with 00014 SV:return code 107 00015 ST:VMFSET2204I Linking MAINT 4C2 as 4C2 with the link mode MR 00016 SV:HCPLNM107E MAINT 04C2 not linked; not in CP directory 00017 SV:VMFSET1965E The command, CP LINK MAINT 4C2 4C2 MR , failed with 00018 SV:return code 107 00019 ST:VMFSET2204I Linking MAINT 4D2 as 4D2 with the link mode MR 00020 SV:HCPLNM107E MAINT 04D2 not linked; not in CP directory 00021 SV:VMFSET1965E The command, CP LINK MAINT 4D2 4D2 MR , failed with 00022 SV:return code 107 00023 ST:VMFSET2204I Linking MAINT 4A6 as 4A6 with the link mode MR 00024 SV:HCPLNM107E MAINT 04A6 not linked; not in CP directory 00025 SV:VMFSET1965E The command, CP LINK MAINT 4A6 4A6 MR , failed with 00026 SV:return code 107 00027 ST:VMFSET2204I Linking MAINT 4A4 as 4A4 with the link mode MR 00028 SV:HCPLNM107E MAINT 04A4 not linked; not in CP directory 00029 SV:VMFSET1965E The command, CP LINK MAINT 4A4 4A4 MR , failed with 00030 SV:return code 107 00031 ST:VMFSET2204I Linking MAINT 4A2 as 4A2 with the link mode MR 00032 SV:HCPLNM107E MAINT 04A2 not linked; not in CP directory 00033 SV:VMFSET1965E The command, CP LINK MAINT 4A2 4A2 MR , failed with 00034 SV:return code 107 00035 ST:VMFSET2204I Linking MAINT 4B2 as 4B2 with the link mode MR 00036 SV:HCPLNM107E MAINT 04B2 not linked; not in CP directory 00037 SV:VMFSET1965E The command, CP LINK MAINT 4B2 4B2 MR , failed with 00038 SV:return code 107 00039 ST:VMFSET2760I VMFSETUP processing completed unsuccessfully 00040 SV:VMFSUI1965E The command, VMFSETUP, failed with return code 100 when 00041 SV:issued with the argument(s): SERVP2P LE (LINK NOPROMPT 00042 SV:RETAIN D 00043 ST:VMFSET2760I VMFSETUP processing started for DETACH LE 00044 ST:VMFSET2760I VMFSETUP processing completed successfully 00045 ST:VMFSUI2760I VMFSUFIN processing completed unsuccessfully 00046 ST:VMFSUI1211I An Initial Restart Record has been created for package 00047 ST:RPTF0168 in the System-Level Restart Table 00048 SV:VMFSRV1965E The command, VMFSUFIN, failed with return code 100 when 00049 SV:issued with the argument(s): ALL ( NOPROMPT RSUENV 00050 SV:RPTF0168 00051 WN:VMFSRV2310W Service restart file, SERVICE $RESTART A, has been created 00052 WN:due to errors. Correct the errors, and restart SERVICE 00053 WN:using the following command: 00054 WN:SERVICE RESTART RPTF0168 00055 ST:VMFSRV2760I SERVICE processing completed unsuccessfully Someone please say Adam, you're an idiot, you obviously forgot the somethingorother and not Oh! Wow. Yeah, you're right! The RSU doesn't actually work on filepool systems. Reinstalling the system from scratch--while not technically challenging--*will* tick me off. Adam
Re: Please tell me I did something stupid
On Mar 17, 2009, at 2:43 PM, Neale Ferguson wrote: The component is cpsfs or cmssfs ... Not cp or cms which are for minidisks. I just tried to run SERVICE ALL RPTF0168 Which, you know, *should* be able to figure it out. It has in the past. Adam
Re: Please tell me I did something stupid
On Mar 17, 2009, at 2:53 PM, O'Brien, Dennis L wrote: Adam, It looks like SERVICE is trying to service LE instead of LESFS. There are a couple of items in the PSP bucket that have to do with SFS, but the descriptions aren't an exact match for this problem. Did you read all of the ZVM540 subsets and do what they said? All I did was download the RSU and then follow the Service Procedure paragraph on the Quick Install Guide. U. I didn't run instvm dvd after doing the installation. Hmmm. Let's see if that does the trick. The messages look very promising. yeah, that was it. I did in fact do something stupid, and that was to not actually run instvm dvd after the DVD install. Thanks! Service is proceeding now. Adam
Re: Please tell me I did something stupid
On Mar 17, 2009, at 5:03 PM, Alan Altmark wrote: On Tuesday, 03/17/2009 at 04:11 EDT, Adam Thornton athorn...@sinenomine.net wrote: All I did was download the RSU and then follow the Service Procedure paragraph on the Quick Install Guide. Because there's only so much you can put on one page and have it remain legible, the Quick Guides are going away. As this illustrates, they have outlived their usefulness and now pose a clear and present danger to society. (Please don't leave a copy of a Quick Install Guide in a crib - choking hazard.) BOO! HISS!!! I will *read* a one page guide. Although I will apparently SKIP WHOLE STEPS, like INSTVM DVD. a 300 page manual? Not so much. So, any idea why the DOSINST and CMSDOS segments didn't want to rebuild (I stupidly had CF1 accessed and was screwing around with SYSTEM CONFIG while PUT2PROD was running, so I had to rebuild CMS and all the segments the old-fashioned way)? I seem to recall this failing before sometime. I don't think I've ever actually used those segments, so I doubt it matters. Adam
Re: Total VDISK space allowed
On Mar 13, 2009, at 10:54 AM, Martin, Terry R. (LOCKHEED MARTIN Performance Engineering/CTR) (CTR) wrote: Rob, I am setting these disks up using SWAPGEN. I am adding enough to total about 4G(This is about what my paging subsystem will handle). This is the test that we spoke of! SWAPGEN (REUSE will let you use VDISK defined in the directory, if you weren't already aware of that parameter. Adam
Re: Total VDISK space allowed
On Mar 13, 2009, at 1:06 PM, Martin, Terry R. (LOCKHEED MARTIN Performance Engineering/CTR) (CTR) wrote: Hi Adam, What is the format of the SWAPGEN command if you use it in the directory I currently issue SWAPGEN in the PROFILE of the z/Linux guest. You use SWAPGEN in PROFILE. It's just that you can use it to format VDISKS specified in the directory. For instance, if you had a VDISK at 160 defined in your directory, you could do SWAPGEN 160 nblocks ( REUSE Where nblocks is however many blocks. (The next version probably should figure out how big the device is and default to the biggest possible size in the absence of nblocks, huh?) Adam
Re: Total VDISK space allowed
On Mar 13, 2009, at 3:47 PM, Rich Smrcina wrote: I was one of the users that requested this. The advantage is that the size of the VDISKs are then controlled by the directory entry. You don't need to have special handling in the PROFILE EXEC for each virtual machine if a different size is required for whatever reason. Martin, Terry R. (LOCKHEED MARTIN Performance Engineering/CTR) (CTR) wrote: Hi Adam, Thanks for the information. What advantage does defining the VDISK in the directory and using SWAPGEN in the PROFILE with REUSE over not defining them in the directory and just doing the SWAPGEN in the profile without REUSE? Just trying to make sure I understand this! What Rich said: it can give you better control over who gets how much swap space. Although, er, you DO need special handling in PROFILE because the number of blocks can't default. The usual case for it is: I only want most users to be able to request 50MB of VDISK, but I have one large Linux guest that needs 300MB. You can set syslim and userlim appropriately and just give that one guest a giant VDISK in his directory. Adam
Re: SHARE in Austin
On Feb 28, 2009, at 2:02 PM, Rick Troth wrote: On Thu, 26 Feb 2009, Jim Bohnsack wrote: There should be a SHARE Economic Stimulus Plan for the poor companies (and universities) who are cutting their budgets. I live just north of Dallas (Plano) and offered to pay my own transportation, meals, and Motel 6 charges but the $1800 or so for the SHARE registration is a show-stopper, according to my manager. I still work for Cornell--I'm just a longer walk to the computer room. I'll be sticking my nose out the door to catch the southern breezes that come up from Austin next week. Jim Somehow southern breezes and Austin just don't paint that pretty of a picture, Jim. I know you wanna be there, and I don't blame you, but ... eeeww. It is still an improvement over Plano's natural scent. Adam
Re: z/VM 5.4.0 RSU 801 - System Abend Code - VAI008
On Feb 25, 2009, at 2:58 PM, Raymond Noal wrote: Dear list, Does anyone know what this system abend code is – VAI008? The CP Codes manual only says that the abend code is issued by module HCPVAI and that this module is object code only with no source available. The codes manual makes no attempt to give a description of what caused the abend. Steve Vai broke a string. Is it Friday yet? Adam
Re: SSLSERV question
On Feb 17, 2009, at 4:21 PM, Alan Altmark wrote: On Tuesday, 02/17/2009 at 01:41 EST, clifford jackson cliffordjackson...@msn.com wrote: I am in the process of building a SSLSERV virtual machine, under z/VM 5.3 SLU 801, using SLES 9 SP3. ONE question is there a Red book for this process?.. No. We tried to make the instructions in the TCP/IP Planning Admin book as easy to follow as we could. (There is a paper floating around Somewhere out there written by some IBMers, but I wouldn't use it.) Or you (meaning Clifford, not Alan) could save yourself the pain, and go to http://www.sinenomine.net/products/vm/sslenabler and follow the instructions thereon. Adam
Re: zvm 5.2 storage limits, zlinux?
On Jan 21, 2009, at 2:34 PM, Dean, David (I/S) wrote: We have a virtual zLinux SUSE 10.1 file server on zVM 5.2 that currently supplies over 600 Gigs of DASD (SAMBA 3 file storage) to end users. Is there a point (technical or logical) when we should build a second server rather than continuing to grow this server? Probably. But 600GB isn't it. Well, I guess ext3 has an 8TB volume limit at 4K blocks. ext4 removes this, I'm pretty sure. Adam
Re: Private Subnet for Hipersocket connections
On Jan 14, 2009, at 10:43 AM, David Boyes wrote: This is also why Sun stopped using real addresses in their documentation examples. Too many people actually set their systems up to run using Sun's actual address space and when they connected to the public Internet, Extremely Weird Things happened, followed by mass renumbering projects. I don't care who you are: you're vanishingly unlikely to be using ALL of net 10/8; that's 16777214 addresses. 172.16/12 has 1048574 available addresses. 192.168/16 only has 65534. Assuming you're not subnetting. But even so, you have quite a lot of address space in those three ranges. Adam
Re: [IP] Creating a rogue CA certificate
On Dec 31, 2008, at 4:34 PM, David Boyes wrote: Interesting new attack on SSL-based security - compromise the CA infrastructure. Amazing that CAs still use MD5. I would have expected Thawte, anyway, to know better. Adam
Re: TAFOT: *country*?
On Dec 18, 2008, at 8:51 AM, David Boyes wrote: On 12/17/08 9:09 PM, Alan Altmark alan_altm...@us.ibm.com wrote: Thank you for running the test. As Garth Brooks says, Thank God for unanswered prayers. :-) Chuckie listens to *country*? Much becomes clearer. 8-) Do they have both kinds of music there in Endicott? Actually, if it's Garth Brooks, it's more like country. It might be hard to see from there, but those *are* sneer quotes. Garth is the Anti-Hank, as Kinky Friedman likes to say, and that kind of country ain't nothing but pop in a cowboy hat and tight jeans. Adam
Re: Starting an exec on a remote machine
On Dec 4, 2008, at 7:32 PM, Marcy Cortes wrote: Personally, I'd like that product to work hand-in-hand with our existing VM:Backup product which we z/VM'ers control on our own. (Anyone at CA listening) Yes, what he said! If anyone on the list is interested in building a solution around Bacula on your Linux guests talking to VM:Backup (or DFSMS) on the back end and using its catalog, well, you know where to find us. Adam
Re: Web servers for VM
On Dec 2, 2008, at 11:17 AM, Dave Jones wrote: Another limitation is that WEBSHARE does not support SSL, just plain HTTP. The RSK-based one from IBM doesn't either. Yeah, but it'd be trivial to wrap it in SSLSERV. It's a well-behaved protocol, not like FTP. Adam
Re: LOGONBY - limit of 8 userids.
On Nov 25, 2008, at 9:43 AM, Alan Altmark wrote: 2) they must change as CP changes. [free advice: changes are a- comin', rollin' 'round the bend.] Should we be using Folsom Prison Blues or Dylan's Slow Train as our model? Adam
Re: Sharing PPRC devices with z/OS
On Nov 25, 2008, at 10:38 AM, Alan Altmark wrote: Oh, z/OS! z/OS! Another drink, if you please. Be sure to put an umbrella in it. A blue one. No no, not sky blue, but one of azure. You're such a dear. Now go on and play - I'll call again if I need you. This is disturbing for very many reasons. Adam
Re: Number os SSL connections
On Nov 17, 2008, at 10:32 PM, David Boyes wrote: You can try. 8-) The code up to and including the 5.3 release won't go much over 192, even with current service. We have a version of the SSL Enabler appliance that reflects the latest IBM code if you want to try it. Note: latest == latest 31-bit, which is not the full-on 2000-users patch mentioned earlier. The Sine Nomine SSL Enabler is 31-bit only. I have not yet rebuilt it for 64-bit. The non-zSeries alternative does not have any such limitation, however. Adam
Re: Header file to COBOL copybook?
On Nov 13, 2008, at 7:53 AM, Phil Smith III wrote: Anyone have any tools for converting C header (H) files to COBOL copybook files? Or experience doing so? Does a bottle of Scotch and another bottle of aspirin count as a tool? (MicroFocus COBOL for Unix had a tool to do this: H2cpy) Adam
Re: SWAPGEN not working for FBA
On Nov 7, 2008, at 4:30 PM, Martin, Terry R. (CMS/CTR) (CTR) wrote: Hi I am trying to use the SWAPGEN EXEC to create a FBA SWAP file by adding it to the PROFILE EXEC of the Linux guest. However it seems to keep taking the default of DIAG. Since I do not have the DIAG drivers on my RedHat REL4 guest I need to do FBA. Here is he syntax I am using: 'SWAPGEN 900 524288 FBA' I am tracing the EXEC now to see if I spot something. BTW I do have the RXDASD MODULE on the guest. This module is required for FBA swap. Does anyone have any ideas about this? SWAPGEN 900 524288 ( FBA It's an option. It comes after an open-paren. However, I think the FBA driver will actually work with a CMS- formatted disk and partition 1 of the device (rather than the raw device) if you leave DIAG off. Adam
Re: Slow SSH response
On Nov 5, 2008, at 8:19 AM, Steve Mitchell wrote: The puzzle: Why did the SSH process fail? I'm not certain where to look for an explanation. I've checked the Velocity reports, VM was doing some paging during both of these times, could that have done it? CPU consumption was not excessive at approx 50%. Where else might I look? Do the guests have correctly configured nameservers? I've seen ssh be really slow while it tries to do a lookup on the connecting IP if there's a nameserver timeout. Adam
Re: Recycle yourself
On Oct 30, 2008, at 10:44 AM, Schuh, Richard wrote: I would like to see how that would be implemented, the die and come back part, without some external agent being involved. This is getting way too theological for me. Adam
Re: Reliability of SFS?
On Oct 29, 2008, at 12:56 PM, Alan Altmark wrote: SFS has been around for 21 years No! That's impossible! Why, that would mean that I'moh, dear. Adam
Re: Linux guest 191/200 disk question
On Oct 28, 2008, at 12:32 PM, Tom Duerbusch wrote: 1. As has been said, you don't need a R/W disk to IPL. R/O is good. SFS directory is even better. 2. Once you IPL Linux, you are not in CMS anymore. You won't be doing anything with your a-disk anymore. So make it easy on your self, when you need to make changes to the profile exec. Put it in a SFS directory. And then export SFS via NFS? Linux doesn't speak SFS either. With minidisks you can use cmsfs to read what's on them. A port of IPGATE to Linux would be sort of cool, but way more effort than just export SFS via NFS. Adam
Re: Linux guest 191/200 disk question
On Oct 28, 2008, at 1:36 PM, Tom Duerbusch wrote: I must of missed the first part of the conversation Why would you want Linux to have access to your A-disk? There might be reasons, but inquiring minds want to know, and deleted the original posts G. Handy for building systems where you can change Linux behavior without the user knowing much of anything about Linux, by editing files in CMS. Adam
Re: TN3270 for Apple's iPhone....
On Oct 24, 2008, at 9:11 AM, Dave Jones wrote: Found over on the IBM-MAIN list.looks interesting: http://www.mochasoft.dk/iphone_tn3270.htm David's already using it. He says it's pretty good. Me, I'm getting a G1. I crushed my MDA when I fell on it, and T- Mobile has been OK so far. Should be here in a couple weeks Adam
Re: Redbook or Whitepaper For Deploying SSL on z/VM
On Oct 24, 2008, at 10:12 AM, Michael Coffin wrote: Hi Folks, Does anyone know if there is a Redbook, Whitepaper or other good reading for deploying SSL on z/VM for the first time? I'm trying to follow the instructions in the TCP/IP Planning Guide, but I have to confess as a first-timer they are a bit overwhelming. Thanks in advance for any suggestions. Well, the fact that it's a giant pain is sort of the reason we developed the SSL Enabler. The bad news is that this depends on the version of z/VM you're running. The Enabler itself is free; we'll sell you ongoing support for it if you would like; contact me offline if you're interested. If you're running z/VM 5.3, http://www.sinenomine.net/products/vm/sslenabler is what you want. You will need to provide us proof of your current VM license. If you're running 3.1-5.2, the licensing stuff is the same but I need to get you a different download which is not currently linked on the web site. Talk to me offline. If you're running 5.4, it's no longer Linux-based and thus the Enabler won't help you. Adam
Re: z/VM JAVA VM
On Oct 22, 2008, at 3:04 PM, David Boyes wrote: times earlier, if you want JAVA support on VM, you should install a Linux guest and use the up-to-date levels on there. Java itself is a virtual machine. Maybe a z/Java guest someday? Would be a clever way to actually make the zAAP specialty engine useful to a z/VM system. Someone would have to write a Java operating system, and I don't know if the zAAPs can actually do I/O. Still, SET MACHINE JAVA would be interesting. http://www.jnode.org/ Of course, it seems to assume that the thing you're running it on is x86. Not that that's so bad if the assembler nano-kernel really is small, but the graphical console and VESA support don't make a lot of sense on z. If one, however, is speculating about a pluggable backplane with various engine types on it, well, x86 would certainly be one of those, but you could also skip the middleman and run a Java processor. http://www.jopdesign.com/ would be one such, although that's FPGA rather than real dedicated hardware. Still Adam
Re: Question
On Sep 30, 2008, at 1:13 PM, Martin, Terry R. (CMS/CTR) (CTR) wrote: As a matter of fact that is one of the commonalties of these hosts. There is Oracle Clustering going on in these hosts. Can you explain more about this? Well, I don't know exactly how Oracle does it. But I do know, for instance, that for Veritas clustering, you need to define a set of shared disks on a controller that implements SCSI reserve-release, and that Veritas uses that to see who's actually active in the cluster. They recommend that you use the tiniest LUNs you can, because they basically just write a few bytes to the start of the physical device and those bytes differ depending on who wrote it. So I wouldn't be surprised if the following were true: Oracle does something similar (and certainly you CAN do something like reserve/ release with channel-attached DASD), but owing to miscommunication, you're actually putting filesystems and data on the disks you were supposed to be using just for cluster arbitration management. Then, every so often, one of the cluster members stomps on part of your disk, and everything falls apart. Mind you, this is total guesswork. It might be something else entirely. Adam
Re: Question
On Sep 29, 2008, at 3:54 PM, Martin, Terry R. (CMS/CTR) (CTR) wrote: Hi Over the last month or so we have had CHECK SUM ERRORS on 3 of our z/ Linux hosts. This error stops the Linux host from coming back up after a re-boot or log off. After working with REDHAT they found that there was 2 bit over lay of what amounts to the VTOC which points to the UUIDs. Each time this has happened it has been the same over lay. The common thing on these hosts are that they all run Oracle 10g, REDHAT REL4, and FDR/UPSTREAM. When this happens we must boot in RESCUE mode and re-build the UUIDs (not sure of this process by Linux guy does this). I was just wondering if anyone has seen this type of issue. This is our POC but if this does not get resolved we will be hard pressed to move forward. Those disks don't participate in some sort of Oracle clustering arbitration scheme, do they? That sort of very low level overwriting of the disk is the sort of thing I'd expect to see in something was going wrong with a cluster filesystem that used the platters to do who's-got-the-rock negotiation. Adam
Re: WAIT STATE
On Sep 14, 2008, at 3:33 PM, Martin, Terry R. (CMS/CTR) (CTR) wrote: Hi I am having a little problem maybe you can see what I am doing wrong! I defined the mdisk starting at 38 ending at 159 for my parmfile: q v 195 DASD 0195 3390 53DRES R/W159 CYL ON DASD 514B SUBCHANNEL = 000E When I trying accessing it I get this: acc 195 x DMSACP112S X(195) device error Any idea what I am doing wrong? Is it CMS formatted? Adam
Re: question to backup of osa-icc settings
On Sep 1, 2008, at 7:04 PM, David Kreuter wrote: HMC worksheet? Who fills that in? What's wrong with a napkin? This one has mustard on it. At least, I *hope* it's mustard. Adam
Re: VM size for a 2nd level VM
On Aug 28, 2008, at 10:54 AM, Duane Weaver wrote: Well here is the scoop. We acting as a DR site for another university. The other university wants to bring in their zVM 5.3 and run it under our zVM 5.2 system. Our z800 is running in basic mode with 1 lpar, running the zVM 5.2. Shouldn't the sizing depend on the size of the VM system they want to bring in? I mean, if they're bringing in a 64GB system and you've only got 16GBsomeone's gonna be unhappy. Adam
Re: Request for information: Installation Summary Cards
On Aug 28, 2008, at 8:07 PM, Alan Altmark wrote: The Court of Opinion and Assizes is now in Session. Recall that we have two one-sheet (two pages, front back) tri-fold installation summary cards: one for tape and one for DVD (1st and 2nd level). The questions: 1. Do these cards have value? 2. If they are valuable, are they usable in their current form? 3. If you could make changes to them, what would they be? 1 2: YES! Seriously: VM has the BEST INSTALLATION MANUAL EVER. I absolutely ADORE the simplicity of it. DON'T MESS WITH IT. So, 3: There's not a lot I would change. Seriously: it tells you what you need to get the system installed, and it FITS ON A SINGLE SHEET OF PAPER. Adam
Re: Where Do I Go From Here?
On Thu, Aug 21, 2008 at 08:19:22PM -0400, David Boyes wrote: If you truly still need 370 mode, you're going to be hard pressed to find a system that will still run true 370 mode. Most (if not all) the modern systems no longer have true 370 mode microcode. Howeveron a modern z9 you can probably run Hercules and get as much speed as you ever would have on a real 370. Hercules can emulate a 370 very accurately, and if you can license the system to your actual CPU, then it might even be legal. I doubt anything specifies WHAT the intervening virtual machine layers are. Adam
Re: DDR'ing 3390 DASD To Remote Location
On Aug 19, 2008, at 1:39 PM, Fran Hensler wrote: On Tue, 19 Aug 2008 13:21:40 -0400 Jiri Stehlik said: http://www.vm.ibm.com/download/packages/descript.cgi?DRPC The FTPPUT and FTPGET PIPE stages were also included and documented at the above address. I have the latest CMS PIPLINES but it doesn't include FTPGET and FTPPUT. I can't find them on the IBM Download site either. Where can I get them? Run INSTPIPE MODULE from MAINT 2CC. Adam
Re: DDR'ing 3390 DASD To Remote Location
On Aug 19, 2008, at 1:48 PM, David Boyes wrote: Well, far be it from me that I suggest that VM Development begin to talk to themselves. You lot 're odd enough to begin with...8-) As Zork so eloquently put it, Talking to yourself is said to be a sign of impending mental collapse. Adam
Re: DDR'ing 3390 DASD To Remote Location
On Aug 19, 2008, at 2:29 PM, Fran Hensler wrote: On Aug 19, 2008, at 1:39 PM, Fran Hensler wrote: I have the latest CMS PIPLINES but it doesn't include FTPGET and FTPPUT. I can't find them on the IBM Download site either. Where can I get them? On Tue, 19 Aug 2008 13:48:20 -0500 Adam Thornton said: Run INSTPIPE MODULE from MAINT 2CC. I'm stuck on z/VM 3.1 because I am on a FLEX-ES box. I found INST* files on MAINT 2C2 but not FTPGET or FTPPUT. I have the latest Princeton Runtime Distribution but there is no FTPGET or FTPPUT stages. Fortunately, then, they are in the DDR stage that George just added. The stages appeared in 5.1 to support the DVD install of z/VM. I would not expect earlier releases to have them on MAINT 2CC. Adam
Re: Linux Commands
On Aug 15, 2008, at 8:58 AM, Rich Greenberg wrote: On: Thu, Aug 14, 2008 at 10:03:14PM -0500,Adam Thornton Wrote: } On Aug 14, 2008, at 9:34 PM, Paul Raulerson wrote: } } (1) Learn vi. } } Heretic. Adam may call me a heretic also but I agree with Paul. While there are many other editors available on *ix and you can easily start a holy war over which is the best, vi is the editor that you can be sure is there. Want to install EMACS (aka Eight Megs And Constantly Swapping) or THE or any others on a newly installed *ix? Use vi to tailor the makefile and config files first. Coming in as a consultant and the system doesn't have SuperEditorX which you are used to, and you don't have the time to learn SuperEditorY, it will almost certainly have vi. Rich is, I hate to say, right. Very few systems will not have vi preinstalled. Emacs is hardly ever already-just-there (yay OS X!). So you SHOULD learn enough vi to get by. Fortunately, that's pretty much: hjkl, 0 and $, escape, /, dd, x, i, a, w, q, and ! Adam
Re: Linux Commands
On Aug 14, 2008, at 7:08 PM, Thomas Kern wrote: When CMS HELP first came out, the group I was with built a process to format and print all of the Help files into our own books. It would be nice if there was a process to format and print all of the MAN pages that are resident on an arbitrary linux system (z or x86). for i in $(find /usr/share/man/man*) -name \*gz; do gzip -dc $i | nroff | lpr done No, don't actually do this. You will be sorry. Adam
Re: Linux Commands
On Aug 14, 2008, at 9:34 PM, Paul Raulerson wrote: (1) Learn vi. Heretic. Adam
Re: Safety Reminder: If you are planning disk upgrades, make sure you switch your Linux guests to by-path IDs in /etc/fstab BEFORE you switch
On Aug 12, 2008, at 11:47 AM, Stephen Frazier wrote: Two questions that anyone who is new enough to need your reminder will ask are: What default do you suggest? When changing it, how should it be changed? The old timers here will know the answers. *I* suggest using the /dev/disk/by-path filename. (I think RH is /dev/ dasd/by-path). Change it with your favorite text editor in /etc/fstab *and* in /etc/ zipl.conf, and don't forget to rerun zipl! Adam
Re: Safety Reminder: If you are planning disk upgrades, make sure you switch your Linux guests to by-path IDs in /etc/fstab BEFORE you switch
On Aug 12, 2008, at 12:25 PM, Thomas Kern wrote: I have been using /dev/dasd?1 where ? goes from a to zz. Is by-path the /dev/disk/0.0.0591 syntax? Yeah, although it's more like /dev/disk/by-path/ccw-0.0.0591-part1 these days. Ada,
Re: OT: Alan has a pony.
On Aug 12, 2008, at 4:31 PM, David Boyes wrote: I would just like to point out that Alan Altmark’s long-standing wish for a pony has been satisfied. A brown and white pony has been delivered, and he has no need for further ponies. 8-) Uh oh. See, it wasn't Alan who wanted the pony. It was Chucky. That poor, poor pony. Adam
Re: SWAPGEN EXEC
On Aug 7, 2008, at 4:28 AM, Berry van Sleeuwen wrote: Hello List, The past day I have tried to get the new version of SWAPGEN into my VM but no success so far. I have tried every way I can do a filetransfer but every time I end up in a file I can't use on VM. Can anyone send me the VMARC-ed version of the current SWAPGEN so I can continue? I do have the old version here but there seemds to be a problem with my VDISK swap so I'd like to try the newer version. (TIA) It's unlikely to helpthe recent changes are pretty minor. Nevertheless, the VMARC of the new version is now available at: http://download.sinenomine.net/swapgen/ BTW, when I look at the mailable I see the filelist of the included files. The RXDASD MODULE in mailable archive looks like to be from 3/15/04. I too have a RXDASD but that one is dated 18/11/95 and that is also the one from the IBM VM downloads. Is this RXDASD a different version or are they the same? Probably the same. I just wasn't careful about preserving the timestamp, but it just comes from the VM Downloads page. I've looked into the comments made here in June, and I have to agree with the comments Adam made (to go back to the VMARC format). IIRC, VMARC was recommended to use because of the fact that ASCII to EBCDIC translations vary among the various file tranfer methods. (Either FTP or IND$.) So by replacing VMARC with MAILABLE to avoid the need for some extra tool and/or to overcome less than default record formats, the transfer now relies on the fact of how your method of transferring files will handle ASCII to EBCDIC translation. IMHO this is much harder to solve than a faulty recordformat because the file itself will be altered by the transfer. At least with VMARC or CMS packed I know what recordformat it should be (either Fixed 80 or 1024) and can FBLOCK it after transfer. And because these files must be transferred in binary the file itself should not change by the transferprocess. Regards, Berry. Also note that Leland Lucius has written VMA, which is a portable utility to do vmarc manipulation. http://www.homerow.net/zvm/vma.htm This is handy if, for instance, you just want to extract the docs from a package and read them on the desktop system of your choice. Adam
Re: SSL connection problem after IPL
On Aug 6, 2008, at 7:37 AM, Tim Joyce wrote: Hey guys, I've been using secure telnet through a SSLSERV for many months now. After we IPLed over the weekend, it stopped working. The SSLSERV machine is up a communicating with SSLADMIN commands. The certificate looks correct. NETSTAT CO shows the SSLSERV in a listen status. But, if I try to start up a secure telnet session, I get : SSLv2 handshake failure: Socket error 1: SSL_ERROR_SSL. Can anyone point me in the right direction? Log into the console of the SSLSERV machine. Run df and see if maybe your log files have filled up the whole disk. Also check your PROFILE TCPIP and make sure that you really are wrapping the ports you think you're wrapping. Adam
Re: SSL connection problem after IPL
On Aug 6, 2008, at 9:35 AM, Tim Joyce wrote: Hey Adam, Thanks for the reply. Here is my DF command: df Filesystem 1K-blocks Used Available Use% Mounted on /dev/dasda1 139368127316 12052 92% / tmpfs63040 0 63040 0% /dev/shm /dev/dasdb165632 592 6% /opt/vmssl/ parms Is 92 % ok? How should I clean up the log files? 92% is fine. Judging from the partition size, you're running z/VM 5.2 or earlier, right? As far as PROFILE TCPIP errors, I did notice yesterday I had misspelled the PORT statement for my SSLSERV admin : TCP SSLSERV SERCUR ALCERT ; SSL SERVER - ADMINISTRATION so I corrected with obeyfile : TCP SSLSERV SECURE ALCERT ; SSL SERVER - ADMINISTRATION If this is the problem, I do not understand why it would have worked before the IPL. And, if this was the issue, shouldn't the corrected obeyfile have resolved this, or will I need to wait until I can cycle the TCPIP stack this weekend? If it worked before IPL it was probably that someone had done an OBEYFILE last time, but I would think an OBEYFILE would have worked this time. How about the ports that you're actually using to connect SSL services on? What do those look like? Do they have the right certificate names? Adam
Re: Nice idea in blog: Should we toss x86 architecture - NOT.
On Jul 31, 2008, at 5:08 PM, Gary M. Dennis wrote: z/VOS translates guest OS code during initial execution. Code fragment storage, lookup, disposal and reuse for primary and sibling guests are addressed in a patent application. Suffice it to say that we don't interpret or emulate massive amounts of x86 code for use 2-n. Ah, the old ARDI Executor approach. Well, probably not *exactly* if there's enough novel in there to patent, but, yeah, JIT instruction-stream-translation with caching of commonly-reused fragments was the way ARDI was emulating 68K Macs on Pentiums back in, oh, 1996 or so. That would (of course) be a better way to cut your overhead, because if you have a long sequence of s390x instructions mapped (nearly 1-to-1) to a long sequence of x86(_64?) instructions, then you only take the wrapping the registers and shifting the arguments around hit once per chunk, and a lot of x86 executable code is a) boilerplate and b) common in most executables produced by the same compiler, so this could actually be a very big win. Does z/VOS do x86_64 or just 32-bit x86? Actually, can you list which x86 extensions are included in it? Or do I need to wait for release to find out. Adam
Re: Loosing IP after IPL: CTC free
On Aug 1, 2008, at 10:09 AM, Rob van der Heij wrote: On Fri, Aug 1, 2008 at 4:40 PM, Alan Altmark [EMAIL PROTECTED] wrote: And while I'm halucinating, I would have a command that adds and deletes users from an autolog list. The list is in the warm start area. No requirement for AUTOLOG1 unless you want it. Oh, and the autologs would be paced to ensure that the system is not brought to its knees during IPL. We did run PROP in AUTOLOG2 and it would also get the message when the autologged user went away, it allowed the operator to check the status, etc. It allowed people to add things to the list (also users could request their own server to be started). IIRC we had a disk with a file per userid to be autologged, and the file itself had information about when it should be started and on what system and who to alert when it broke. I think David recently also did something like this. If by David, you mean Adam, then yes, the answer is we did, it was SYSVINIT. It's at http://download.sinenomine.net/sysvinit/ It wasn't quite as sophisticated as all that, but it's a lot more sophisticated than a list of XAUTOLOGS interspersed with CP SLEEPS. The URL in the WAAV presentation there is no longer valid. But if you got to the WAAV presentation, you're there. The community response to SYSVINIT was deafening silence, which sort of surprised me, as it really DID provide a nice flexible mechanism for automating startup with real dependency checking. Although the log disk fills up--I never got around to adding an autocleaner to it-- and I never got midnight message trapping to work correctly. Still, if there's actually SOME interest in it, I might be eventually persuaded to address that. If not, hey, it's all Rexx, so feel free. Adam
Re: ADD VIRTUAL MEMORY DYNAMICALLY
On Jul 28, 2008, at 2:43 PM, Schuh, Richard wrote: No. Redefining virtual memory causes a virtual system reset. It takes an IPl after that. You can update the directory to allow additional, but you cannot redefine the virtual storage of a running machine without causing the reset. Well, not quite true. You could add more swap. And if you added swap-on-VDISK, it'd really live in memory (assuming no memory pressure on the Linux guest). So, define VDISK to the machine at some address. Bring it online through the /sys/bus/ccw interface. mkswap to format it. swapon /dev/ dasdt1 (or whatever), and you should have a machine able to swap to your new VFB-512 disk. Adam
Re: CSE and VMSERVx
On Jul 25, 2008, at 7:36 AM, Rob van der Heij wrote: Resist the idea to use VMSYS* filepools for other purposes. Any time saved by hijacking the VMSERVU configuration to hold your own data is normally paid back later at a less convenient moment. I consider myself more than usually tolerant of system-administration- related pain, butdamn, SFS doesn't make defining new pools exactly *easy*, does it? Adam
Re: Nice idea in blog: Should we toss x86 architecture - NOT.
On Jul 25, 2008, at 8:48 AM, McKown, John wrote: -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Gary M. Dennis Sent: Thursday, July 24, 2008 9:34 PM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: Nice idea in blog: Should we toss x86 architecture - NOT. z/VOS is written to support the x86 instruction set and the underlying hardware rather than a specific operating system. For example, FreeDos was used as the initial debug target operating system due to source code availability. --. .- .-. -.-- Gary Dennis Mantissa Corporation Somewhat like BOCH? I remember somebody saying that they ran Windows on BOCH on an old P/390. Yeah, that was me. I did indeed boot NT4 on Bochs on a P/390. Or maybe it was an H70: the fact that the Linux box was h1.tx.sinenomine.net makes me suspect that it was Dave Jones' (at the time) H70. http://www.fsf.net/~adam/NT-on-390-desktop.png I went and grabbed the latest Bochs last night. Apparently it now does x86_64 and can run Vista. It will be interesting, once Mantissa is released, to do a speed comparison of it versus Bochs, since both appear to be, essentially, x86 emulators. I haven't been able to build it yet; it, at the very least, requires X libraries, so I need to build a dev box with the appropriate libraries. And of course it's going to bespecialto run it on, ahem, a Flex box running in 64-bit mode on 32-bit (Intel) hardware, which is what we have in-house. Adam
Re: CSE and VMSERVx
On Jul 25, 2008, at 10:40 AM, Schuh, Richard wrote: It doesn't seem that difficult, just follow the instructions. It may be drudgery, but the cookbook does work ;-) OK. Compare to adding more storage to ZFS or adding more storage to a Linux LVM2 volume. Sure, I'm not having to manually push each bit uphill both ways ten miles through a driving blizzard to add it to an SFS volume, butow. Adam
