INFO-VAX Mon, 25 Aug 2008 Volume 2008 : Issue 464 Contents: Re: DEFCON 16 and Hacking OpenVMS RE: DEFCON 16 and Hacking OpenVMS Re: strange tcpip issue Re: strange tcpip issue Re: strange tcpip issue ---------------------------------------------------------------------- Date: Sun, 24 Aug 2008 11:04:29 -0700 (PDT) From: jferraro Subject: Re: DEFCON 16 and Hacking OpenVMS Message-ID: <8f7d0af3-c42e-4f6b-8f5c-fdb560cfdcdd@x41g2000hsb.googlegroups.com> > More seriously though: VAX 7000s in production? Has anyone looked at > how much they're costing vs more recent kit? Maintenance, power, > cooling and square-footage (?) on kit (including storage?) from that > era won't be cheap; when I last looked, in most environments, moving > to something current whilst staying with VMS (and therefore > introducing relatively little risk) would typically have a very short > payback time, maybe a year or two? How often can you get a payback > time of that length? It can get more interesting if the business is > organised in a way where revenue spend and capital spend come under > separate stovepipes, but even that shouldn't be insurmountable. > It is interesting, I assure you. As it goes, several folks over the years have put together business cases to port the existing COBOL (and other) code to JAVA and the like, with the end goal in mind to eliminate VMS. Costs to do so have been exorbitant and so it has made more sense, so to speak, to put VMS back in the corner and forget about it - and continue on our merry way. Being primarily an HP shop, several of us were sitting around one day as the topic of VMS arose. At that point, I decided to fire up VMS on one of the rx6600s we had sitting around as a simple proof of concept (while I awaited media from HP, I got hold of Wherry's simh docs and brought myself up to speed on the install). Interestingly, I approached our [single remaining] VMS admin and really got very little enthusiasm (which may be typical from the crowd :) ) about the experiment. I currently have VMS running atop HPVM on HPUX 11.31 (yes, its pre-release), and have convinced him to "take a look"... ...but that would most certainly be my end goal.... I'd love to be able to bring an IA box along side the VAX systems we have and let them "parallel" for a while (for lack of better terms) to garner some trust in the newer technology (both for myself and our VMSer). > Any competent HP (or reseller) rep should be able to help you put a > nice business case together, then all you need is a technically > satisfactory plan for the migration and a bit of time (again, > depending on location and needs, HP, reseller, or independent > consultant). This is where we hit the proverbial brick wall. VMS is too easily forgotten about, I surmise. Most probably because it rarely rears its head, but either way when I broach the topic I see a lot of folks who'd rather not disturb the nest. I honestly believe that if HP would ramp up their internal training on the product we'd all be much better off. I'll do my best from my end... Joe ------------------------------ Date: Sun, 24 Aug 2008 21:57:17 +0000 From: "Main, Kerry" Subject: RE: DEFCON 16 and Hacking OpenVMS Message-ID: <9D02E14BC0A2AE43A5D16A4CD8EC5A593ED5DE5F5D@GVW1158EXB.americas.hpqcorp.net> > -----Original Message----- > From: jferraro [mailto:jferraro@gmail.com] > Sent: August 24, 2008 2:04 PM > To: Info-VAX@Mvb.Saic.Com > Subject: Re: DEFCON 16 and Hacking OpenVMS > > > More seriously though: VAX 7000s in production? Has anyone looked at > > how much they're costing vs more recent kit? Maintenance, power, > > cooling and square-footage (?) on kit (including storage?) from that > > era won't be cheap; when I last looked, in most environments, moving > > to something current whilst staying with VMS (and therefore > > introducing relatively little risk) would typically have a very short > > payback time, maybe a year or two? How often can you get a payback > > time of that length? It can get more interesting if the business is > > organised in a way where revenue spend and capital spend come under > > separate stovepipes, but even that shouldn't be insurmountable. > > > It is interesting, I assure you. As it goes, several folks over the > years have put together business cases to port the existing COBOL (and > other) code to JAVA and the like, with the end goal in mind to > eliminate VMS. Costs to do so have been exorbitant and so it has made > more sense, so to speak, to put VMS back in the corner and forget > about it - and continue on our merry way. > Do not get caught up in the "we need to keep up with the Jones" mentality that often drives companies to spend huge sums of IT $'s over many years which in the end is not successful anyway. Yes, there may be a place for Java and/or other new technologies, but there is nothing wrong with continuing to use Cobol or Basic or Fortran because in the end, the business you are supporting really could not care less which language you use. What the business does care about is drastically reducing IT costs (not increasing) and IT focussed on providing real value to the business i.e. providing new functionality which makes them more competitive. Hence, this is why many Cust's choose an upgrade and integrate model vs. a replace everything from scratch strategy. Certainly the upgrade and integrate (e.g. web enable existing applications with connectors created to allow sharing of code/data with other app environments) is typically a small fraction of the costs when compared to a rip & replace everything strategy. > Being primarily an HP shop, several of us were sitting around one day > as the topic of VMS arose. At that point, I decided to fire up VMS on > one of the rx6600s we had sitting around as a simple proof of concept > (while I awaited media from HP, I got hold of Wherry's simh docs and > brought myself up to speed on the install). Interestingly, I > approached our [single remaining] VMS admin and really got very little > enthusiasm (which may be typical from the crowd :) ) about the > experiment. I currently have VMS running atop HPVM on HPUX 11.31 (yes, > its pre-release), and have convinced him to "take a look"... > > ...but that would most certainly be my end goal.... I'd love to be > able to bring an IA box along side the VAX systems we have and let > them "parallel" for a while (for lack of better terms) to garner some > trust in the newer technology (both for myself and our VMSer). > [snip ...] If the VAX application is written mostly in Cobol and/or other higher Level apps, then moving to Integrity is likely a very easy move that might be justified by simply looking at the VAX maint contract and DC power / space savings over a 3-4 years period. With current versions of OpenVMS, if it were required, you would then be able to use Java and other web services technologies for new stuff to integrate with other environments as appropriate while at the same time maintaining your existing code in their current languages. Regards Kerry Main Senior Consultant HP Services Canada Voice: 613-254-8911 Fax: 613-591-4477 kerryDOTmainAThpDOTcom (remove the DOT's and AT) OpenVMS - the secure, multi-site OS that just works. ------------------------------ Date: Sun, 24 Aug 2008 20:21:56 +0000 (UTC) From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: strange tcpip issue Message-ID: In article <8iWrk.47097$E41.38763@text.news.virginmedia.com>, "Tim Wilkinson" writes: > True, normally I would have assigned a static address to devices I class as > "Servers", but as my "VAX" is a simh on my laptop which moves round with me, > it was convienient to use DHCP. I have not tried it, but would guess even > static assigned would not cope with the subnet mask my network people have > supplied for my home network. Why? I used to have an ISDN connection with my own subnet (8 addresses, one for the router, one for the subnet, one broadcast, so 5 for machines). Why couldn't you use one of the addresses on the subnet your network people have provided? Of course, static doesn't mean so static that you can't change it depending on your network. :-) > I find it surprising that DEC never "Got it" as they were at the forefront > of embracing networking. But for some reason seemed very reluctant to pick > up on TCP/IP. I remember running Wollongong to allow terminal server > attached users to connect to some of my microvaxes in the 80/90s. Other kit > had unibus hardware boards from Bridge (Later part of 3com) to allow TCP/IP > connected terminal servers to communicate. Even in that era, we were using > VLSM to allocate our allocated "Class B" network across the UK sites. DEC manufactured DECnet, which was superior. Why should they have pushed something inferior? Non-DEC stuff could speak DECnet as well, so it wasn't clear that TCPIP would win in the end. ------------------------------ Date: Sun, 24 Aug 2008 16:46:05 -0400 From: "Richard B. Gilbert" Subject: Re: strange tcpip issue Message-ID: <48B1C88D.2000901@comcast.net> Phillip Helbig---remove CLOTHES to reply wrote: > In article <8iWrk.47097$E41.38763@text.news.virginmedia.com>, "Tim > Wilkinson" writes: > > >>True, normally I would have assigned a static address to devices I class as >>"Servers", but as my "VAX" is a simh on my laptop which moves round with me, >>it was convienient to use DHCP. I have not tried it, but would guess even >>static assigned would not cope with the subnet mask my network people have >>supplied for my home network. > > > Why? I used to have an ISDN connection with my own subnet (8 > addresses, one for the router, one for the subnet, one broadcast, so 5 > for machines). Why couldn't you use one of the addresses on the subnet > your network people have provided? Of course, static doesn't mean so > static that you can't change it depending on your network. :-) > > >>I find it surprising that DEC never "Got it" as they were at the forefront >>of embracing networking. But for some reason seemed very reluctant to pick >>up on TCP/IP. I remember running Wollongong to allow terminal server >>attached users to connect to some of my microvaxes in the 80/90s. Other kit >>had unibus hardware boards from Bridge (Later part of 3com) to allow TCP/IP >>connected terminal servers to communicate. Even in that era, we were using >>VLSM to allocate our allocated "Class B" network across the UK sites. > > > DEC manufactured DECnet, which was superior. Why should they have > pushed something inferior? Non-DEC stuff could speak DECnet as well, so > it wasn't clear that TCPIP would win in the end. > DECNet Phase IV or Phase V? It was clear almost from day one that DECNet V would not fly! TCP/IP had already conquered the world! NCL and its documentation were enough of a PITA that I installed phase IV rather than V on all of my machines. I'm still running Phase IV whenever two or more of my home DEC systems are powered up. DEC networking never seemed to realize that most people were NOT running all the weird point-to-point protocols that they supported. Phase IV on the LAN, for the WAN/Internet TCP/IP all the way! ------------------------------ Date: 25 Aug 2008 00:34:37 +0200 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOeGER) Subject: Re: strange tcpip issue Message-ID: <48b1fe1d$1@news.langstoeger.at> In article , =?ISO-8859-1?Q?Jan-Erik_S=F6derholm?= writes: >Why not just run a TCPIP Services version that works ? The (not so) funny answer: There is none at all (better go with PSC products) The serious answer; TCPIP V5.3 is the last for VAX, so to get what he wants, he would need to force HP to release a current TCPIP for VAX also. Or to switch his hardware platform (like to Personal Alpha - Charon = $)... -- Peter "EPLAN" LANGSTÖGER Network and OpenVMS system specialist E-mail Peter@LANGSTOeGER.at A-1030 VIENNA AUSTRIA I'm not a pessimist, I'm a realist ------------------------------ End of INFO-VAX 2008.464 ************************