Re: How comments treated by DIRMAINT
Re: How comments treated by DIRMAINTThere is actually a tool that does a fair job of recreating the source directory from the object directory. It actually won't put PROFILEs back in, or INCLUDEs, or comments, but can put all object parts back to their source equivalents. The DIRENT tool (in the VM download packages) does this, though when I last looked, it wasn't up to date. I maintained it locally to provide what was missing. Alas, I no longer have access... The nice thing about this tool for me was for viewing of several reference systems (and for disaster recovery and a few other cases). A frontend exec took a parameter for which system I was curious about, linked to the disk with the desired DRCT space, and displayed in XEDIT whatever source I wanted. DIRENT was most often used on a particular USER, but it had the option to view/rebuild the entire DRCT. Reading the object directory was fast and there was no need to wait for a maybe busy DIRMAINT to return a GET, no spool space taken up, no temporary clutter in my reader, especially, no need to even bring up the second level system; just show me what I want. It is possible to see the first level from the second level, given a RR link back upward. Or sideways. Security and outdatedness (though not sure where it stands now) aside, a nice tool. I agree about no purpose for comments in the object directory. The audience for querying the directory seems small, just system programmers, who can get to comment/informational references some other way. General users on our system never needed to see comments. Oh, well, we had an elaborate frontend to DIRMAINT that provided users the same thing, as part of billing information that they entered when they requested their own minidisks! But this data wasn't stored in the source directory, and querying was built into that frontend. And system disks were never considered or treated like user minidisks. Two tools for two audiences is OK. Now, if IBM were to add a versatile frontend to DIRMAINT, that could be useful. All kinds of metadata COULD be useful for different sites. But I imagine that SFS features and declining use of VM by end-users would reduce the market for such a tool. (Actually, the billing-oriented tool was built before the DIRM SAPI interface, which would have made it easier.) I imagine that an ESM could be a good place to store metadata (some is already). And maybe Accounting packages could be integrated with the ESM to use the metadata. So maybe a vendor could run with this, and IBM can develop what they see as more strategic to VM. On 2/11, RPN01 said... snip Second, someone mentioned comments taking space in the object directory... My impression / hope would be that comments would be stripped from the information before building the object directory, since there is no actual purpose for them there, and there isn't a convenient tool to take an object directory and turn it into a source directory. Are the comments actually left in the object directory? If so, MAINT is one of the worst offenders, leaving in the hundreds of links that it uses during the installation as comments. snip
Re: VM/370 Release 6 Waterloo tape (CIA MODS)
The copy on the Hercules groups is not from the correct year to have these mods on it. /Tom Kern Berry van Sleeuwen wrote: Hello Mark, In the Hercules comunity this tape is available in AWS format. On cbttape.org it can be found in the vm/370-r6 page. Regards, Berry.
Re: Any Rumors?
On Fri, Feb 15, 2008 at 9:11 AM, in message [EMAIL PROTECTED], Huegel, Thomas [EMAIL PROTECTED] wrote: I thought of this thread last night while watching 'Deal or No Deal' (yes I do have a social life) Hmm. Since when did sitting at home watching TV become considered as having a social life? Maybe I'm not such a hermit after all. Mark Post
Re: Any Rumors?
Mark, Hmm. Since when did sitting at home watching TV become considered as having a social life? i thought the same thing, but then it occurred to me that there could be a LOT of people sitting around that same TV ... err, uhmm - socializing! throw in the odd dog and a couple of cats and there could be lots and lots of socializing going on. Or... neighbors, the pizza delivery man, a couple of cell phone conversations on the side - you can do a LOT of socializing while sitting at home watching TV. prg Phillip Gramly Systems Programmer Communications Data Group Champaign, IL
Re: Any Rumors?
Hello! I see your point, sir. But how does one go about requesting an invitation? Even though I now have the bandwidth to do webcasts, I do not. They do not excite me as much as being physically present. I should also mention that currently I am gathering research data to make a point concerning the prospects of teaching people the wonders of the mainframe, and that would certainly include z/VM and everything else included. -- Gregg C Levine [EMAIL PROTECTED] The Force will be with you always. Obi-Wan Kenobi -Original Message- From: The IBM z/VM Operating System [mailto:[EMAIL PROTECTED] On Behalf Of Jim Elliott Sent: Saturday, February 16, 2008 2:32 AM To: IBMVM@LISTSERV.UARK.EDU Subject: Re: Any Rumors? ... Besides my only complaint concerning the road show discussions that came up earlier is that NYC wasn't selected as a host city for one of the events. ... Actually there is an event in NYC on 2/26, but it by invitation only. There is a webcast on 2/26 for which a separate invite was sent out. Go to http://www.on24.com/clients/ibm/102818 to register. Jim
Re: z/OS 1.7 under VM 4.4
1) z/OS 1.7 runs as VM guest machine under z/VM 4.4. 2) 2066 does not support ZAAP procesors (the processors for Java code). 3) So, z/VM 4.4 does not emulate ZAAP processors. 4) z/OS 1.7 comes with HCD version that does not support 2066. So, be sure to maintain IOCDS with HCD from z/OS 1.5 or lower. Hope this help. Luis Ordonez At 03:25 PM 2/15/2008, you wrote: Could someone tell me if z/OS 1.7 will run under z/VM 4.4 on a 2066? I know it is unsupported, but one step at a time. Thank you, Scott R Wandschneider Senior Systems Programmer Infocrossing, a WIPRO Company 11707 Miracle Hills Dr. Omaha, NE 68154 Office 402.963.8905
Re: Installation Verification Procedures
Thats a good list. Here are some more ideas from our shop: I'd first want to make sure that VM:Secure works. Trying to log on is a good test. New functions will already have been tested second level. Then I'd want to make sure that we can IPL z/OS and z/Linux guests. Then the rest of the VM:Manager suite. Our current management asks us to test ALL our program products, using th e IVP or whatever other test we can come up with for each one. That exercises CMS. I don't thing that's really necessary, as CMS hasn't changed in a long time and problems with program products are now extremely rare. We also have to test our CP exits and the two mods VMSI VPARS and VTAPE. We test those all second level before we ever get into production, though . We have a couple of times been bitten by something that worked differentl y at first level than second level. In general, I think you should concentrate on whatever is most important to your shop - your bread and butter. So an IBM IVP would probably be of limited use. In our shop, most of the important things, like z/OS guests on some systems, and VM:Sched jobs on others, get run automatically when we bring the systems up, so all we really have to do is watch for a few hours on Sunday. (We IPL very early on Sunday.) We have had a number of problems with DB2 Private Data Spaces in z/VM 5.2.0 and 5.3.0, so that also we test at second level, then again first level. (We still have 5.2.0 running on one last system, due to a problem in this area that we discovered second level. It took us 2 months to reproduce it.) I guess the general principle is that if you have had a problem in some area in the recent past, its important to put that on your test list. Alan Ackerman Alan (dot) Ackerman (at) Bank of America (dot) com On Fri, 15 Feb 2008 12:49:56 -0500, Doug Breneman [EMAIL PROTECTED] wrote: Hi All, Many years ago there was an Installation Verification Procedure for VM. This was mostly directed toward CMS tasks. We found that this was not very useful and removed it from our product documentation. We have a list of some items that we use as an IVP when we get a new release of VM. This is not an official IBM IVP. This include things like: CP IPL (MP) - stop if you cannot do this :-) CMS IPL (190 and Segment) - needed to do many other things. 190 is need ed if the segment fails GCS IPL - needed for RSCS and other applications RSCS Initializes - needed for sending/receiving files PVM Initializes - remote system access SFS Initializes - file storage TCP/IP Initializes Able to logon to the system via telnet - remote system access Able to FTP to and from VM - needed for sending/receiving files NETSTAT command functions - needed for our environment Run test workloads to reach 50-70% CPU load for 4 hours. - load that is reasonable for us to ensure some stability. The test workloads vary. Linux and z/OS each IPL second level. - needed for our environment Workloads must run on both of these guest operating systems. These are some of the things that we do to initially verify the installe d VM. If all of these work, then we have a good chance that we can contin ue to use the new release. The items that are important to you might vary from this list. I hope this list might act as one example. Doug Breneman z/VM Development IBM Endicott, NY Adam Thornton [EMAIL PROTECTED] mine.net To Sent by: The IBM IBMVM@LISTSERV.UARK.EDU z/VM Operating cc System [EMAIL PROTECTED] Subje ct ARK.EDU Re: Installation Verification Procedures 02/15/2008 10:22 AM Please respond to The IBM z/VM Operating System [EMAIL PROTECTED] ARK.EDU
