INFO-VAX Thu, 20 Dec 2007 Volume 2007 : Issue 696 Contents: Re: Apache 2, Multiple Tomcat (5.5.9) instances on OpenVMS Re: Compiling PHP and/or any PHP Extension on VMS Re: ES45 upgrade question Re: ES45 upgrade question Happy DEC-20 Day! IP standby query Re: IP standby query Re: IP standby query Re: IP standby query Re: OpenVMS web site file move to ftp.hp.com server Perl 5.10.0 has been released Re: quick question Re: quick question Re: quick question Re: quick question Re: Volume label. Re: Volume label. Re: Why can't I use an IA64 as a boot server for an Alpha? Re: Why can't I use an IA64 as a boot server for an Alpha? ---------------------------------------------------------------------- Date: Thu, 20 Dec 2007 09:02:30 -0800 (PST) From: bcole@emjmetals.com Subject: Re: Apache 2, Multiple Tomcat (5.5.9) instances on OpenVMS Message-ID: <1cacc3d6-86c0-436b-92a6-d815383504bb@i12g2000prf.googlegroups.com> On Dec 8, 3:32 pm, Arne Vajh=F8j wrote: > bc...@emjmetals.com wrote: > > On Dec 6, 5:41 pm, Arne Vajh=F8j wrote: > >> After thinking a bit about it then I can not come up with a > >> mod_jk2 solution to do that mapping. > > >> It would not work well anyway if you app outputs links that > >> include context path. > > >> If you do not deploy many apps you can define the contexts > >> explicit and specify a path as /dev/app1 . > > > Thanks for the info. I guess I am concerned a bit about switching to > > "mod_jk" from "mod_jk2" since it does not seem to be the default which > > HP pushes when setting up this stuff on OpenVMS. It kind of makes me > > concerned about how up to date the "mod_jk.exe" port is but it looks > > like it should be at least worth trying out. > > Note that mod_jk does not solve the link problem. > > Why don't you use a virtual host on Apache and forward > to two differentTomcatinstances instead of messing with the > path ? > > Arne- Hide quoted text - > > - Show quoted text - Thanks for the suggestions. We did decide to use multiple ports and just have apache forward the calls on to the correct Tomcat instance based on port number. The different paths looked to be too much hassle for too little gain. We appreciate your input. Thanks ------------------------------ Date: Tue, 18 Dec 2007 08:59:37 -0600 From: "Craig A. Berry" Subject: Re: Compiling PHP and/or any PHP Extension on VMS Message-ID: In article <13mdpkqcfd692f6@corp.supernews.com>, Mark Daniel wrote: > Grant Croker wrote: > > Hi, > > > > I have over the last year or so tried to build the PHP source provided > > by HP[1] on OpenVMS. The driver for this is a client request to have > > the PHP Ingres extension[2] built on the same platform. Unfortunately > > I have encountered problems building PHP using the supplied source. What sort of problems? > > Whilst I realize the source code has been provided as-is and "The save > > sets do not include complete build procedures...", I was wondering if > > This statement has always puzzled me. > > Of course there is some value in having access to source if you have a > somewhat academic interest in the way something is implemented but I'm > guessing for the vast majority source is for building, correction > (bugfix), tailoring, extension, improvement ... > > What can be the magic ingredient that they are loath to or cannot share > I wonder? Most open source licenses require making one's changes available, so in the first instance, they are simply complying with the license. The caveats about not supplying the full build procedures may be just caveats and may or may not reflect what they actually supply with the download kits. A quick glance shows some .MMS files under the PHP build directory in the sources, so quite possibly there aren't any magic ingredients missing. If there are things missing, it would more likely have to do with someone having no budget to complete a source release rather than any reluctance to share. > > any brave soul had managed to build PHP on their VMS system? > > Dave Jones of OSU has successfully built (earlier?) versions. Perhaps > he used a particular approach he can comment on. > > As PHP is such a widely deployed component of Web infrastructure it > would be beneficial to the community for it to be taken out of HP's > hands with their (still much appreciated but) widely staggered releases > I'm sure they wouldn't mind either. Perhaps someone needs to take the > (probably to begin with fairly onerous) responsibility for porting and > maintaining a VMS baseline for PHP. One that would allow others to > build all sorts of extensions against. A better solution than taking it away from HP would be for HP to take the lead and contribute assertively to the ports their customers need to function in the modern IT world. > This is certainly the case with Jean-François Piéronne's port of Python. > Always up-to-date and bugfixed. You can download and build it > yourself. Add extensions if you want. Mark, if only there were a few more people out there displaying the same level of skill and generosity as you and J-F Piéronne. I agree that volunteer-driven projects often do a better job of staying current than corporation-driven projects. -- Posted via a free Usenet account from http://www.teranews.com ------------------------------ Date: Thu, 20 Dec 2007 03:55:06 -0500 From: "Carl Friedberg" Subject: Re: ES45 upgrade question Message-ID: <890539d90712200055g66b7720fyf6ad4601f348588e@mail.gmail.com> Hi Frank, One thing to check out is the power supplies. Most 1ghz boxes had the lower power style (720?); the 1250 mhz boxes come with bigger, 1020? watt PS. You should have all 3 of them in there. Probably you can get away with less until you crank up the memory; check with your local field service office, they can give reliable answers to your reasonable questions. Carl On Dec 19, 2007 2:34 PM, FrankS wrote: > On Dec 19, 1:52 pm, John Reagan wrote: > > I *think* it is just swapping the CPU modules. Do you have the latest > > SRM firmware? > > I swapped CPU modules in my ES45 last year. I upgraded the SRM firmware > > and swapped the CPU modules only. > > Yeah, the assumption is that the firmware would be updated, and all > other configuration issues were addressed. I can't find an ES45 > document that says you can pull the 1ghz modules and replace them with > 1.25ghz modules. I can't find anything that says you *CAN'T* do that > either, but I'd prefer to have some sort of official blessing on the > operation. > ------------------------------ Date: Thu, 20 Dec 2007 01:36:52 -0800 (PST) From: Rob Subject: Re: ES45 upgrade question Message-ID: <5a4a4f8f-0239-4213-97bf-93d410d77749@w56g2000hsf.googlegroups.com> On Dec 19, 3:17 pm, FrankS wrote: > Does anybody know whether an ES45 68/1000 can be upgraded to the > 68/1250 cpu modules? Is it just a matter of taking out the old cpu > and popping in the new ones, or is there a motherboard change? > > Any pointers to HP/Compaq reference documents would be very much > appreciated. Frank, I'd give the guys at Island Computers a quick call - I'm sure they wouldn't mind helping you. Rob. ------------------------------ Date: 20 Dec 2007 12:48:00 -0500 From: Rich Alderson Subject: Happy DEC-20 Day! Message-ID: With a wish for the success of emulators of the other 36-bit architectures, and the happiness of all! -- Rich Alderson "You get what anybody gets. You get a lifetime." news@alderson.users.panix.com --Death, of the Endless ------------------------------ Date: Thu, 20 Dec 2007 14:10:03 -0000 From: "issinoho" Subject: IP standby query Message-ID: <13mkttv971udk1a@corp.supernews.com> If I have two nodes in a cluster both with the same secondary IP address then one gets allocated the address whilst the other reports it is in STANDBY and only gets allocated the address if 'something' happens to the live node. My question is: what constitutes the 'something'? What needs to happen on Node A such that node B gets the live IP address? Second question: is there any way of stimulating this swap from either of the nodes? ------------------------------ Date: Thu, 20 Dec 2007 09:35:56 -0500 From: "Richard Whalen" Subject: Re: IP standby query Message-ID: "issinoho" wrote in message news:13mkttv971udk1a@corp.supernews.com... > If I have two nodes in a cluster both with the same secondary IP address > then one gets allocated the address whilst the other reports it is in > STANDBY and only gets allocated the address if 'something' happens to the > live node. > > My question is: what constitutes the 'something'? What needs to happen on > Node A such that node B gets the live IP address? > > Second question: is there any way of stimulating this swap from either of > the nodes? > > If the TCP/IP Stack is MultiNet, then the "something" is a lock that the system offering the address currently owns and all other systems that are acting as stanbys are waiting in line to own. You can tell if a system own the lock (and address) with: $ multinet netc clusteralias show You can stimulate a release with: $ multinet netc clusteralias release {ip-address | ALL} ------------------------------ Date: Thu, 20 Dec 2007 14:51:09 -0000 From: "issinoho" Subject: Re: IP standby query Message-ID: <13ml0b0m1pppiba@corp.supernews.com> "Richard Whalen" wrote in message news:fkdugm$9l5$1@news.process.com... > "issinoho" wrote in message > news:13mkttv971udk1a@corp.supernews.com... >> If I have two nodes in a cluster both with the same secondary IP address >> then one gets allocated the address whilst the other reports it is in >> STANDBY and only gets allocated the address if 'something' happens to the >> live node. >> >> My question is: what constitutes the 'something'? What needs to happen on >> Node A such that node B gets the live IP address? >> >> Second question: is there any way of stimulating this swap from either of >> the nodes? >> >> > > If the TCP/IP Stack is MultiNet, then the "something" is a lock that the > system offering the address currently owns and all other systems that are > acting as stanbys are waiting in line to own. You can tell if a system > own the lock (and address) with: > $ multinet netc clusteralias show > You can stimulate a release with: > $ multinet netc clusteralias release {ip-address | ALL} > Sorry (D'oh!) should have said originally. HP TCP/IP Services for OpenVMS Industry Standard 64 Version V5.6 on an HP rx2660 (1.59GHz/6.0MB) running OpenVMS V8.3 ------------------------------ Date: Thu, 20 Dec 2007 07:07:59 -0800 (PST) From: Volker Halle Subject: Re: IP standby query Message-ID: <7a81029c-f931-4422-83ed-4b7fc8096b49@s19g2000prg.googlegroups.com> You may want to read Chapter 5 Configuring and Managing failSAFE IP in the HP TCP/IP Management manual, if your IP stack is TCP/IP Services. http://h71000.www7.hp.com/doc/83final/6526/6526pro_005.html#failsafe_chap FailSAFE IP will monitor the rcv/xmt counters of the active interface and release the IP address, if there seems to be a problem. You can also manually trigger a 'failure' of the active IP interface. Volker. ------------------------------ Date: Thu, 20 Dec 2007 05:36:34 -0800 (PST) From: etmsreec@yahoo.co.uk Subject: Re: OpenVMS web site file move to ftp.hp.com server Message-ID: <1064a1ab-6e3d-482a-9986-7cbef135f840@s19g2000prg.googlegroups.com> On 19 Dec, 19:48, "warren sander" wrote: > I'm in the process of moving all the 'big' files from the OpenVMS web server > to a new home on ftp.hp.com/pub/openvms > I'm doing this for a number of reasons, but for you all it will make the > downloads faster because ftp is a faster more efficient > protocol than http is for large files. > > However, in doing this move I'm finding all sorts of old crufty stuff and > I'm have loads of fun removing it and recoding to point to > the ftp server. > > If you find a download for something is broken please send me an email > warren'dot'sander'at'hp'dot'com and let me know what > page you found the broken link on. > > Most of the files you should be able to find at ftp.hp.com/pub/openvms in a > subdirectory (you should be able to figure out which one and if > not let me also know). > > There are a couple of directories that you won't be able to do directories > of and that is because we don't want you doing directories in > those directories. However there shouldn't be any 'public' software located > there when I am finished. > > -warren Thanks for letting us know Warren. Fingers crossed you'll manage to catch all the old crufty stuff and there won't be any problems. Steve ------------------------------ Date: Thu, 20 Dec 2007 10:27:39 -0600 From: "Craig A. Berry" Subject: Perl 5.10.0 has been released Message-ID: There is a general announcement here: http://news.perlfoundation.org/2007/12/perl_510_now_available.html and I've written up some VMS-specific notes on the vmsperl mailing list: http://www.nntp.perl.org/group/perl.vmsperl/2007/12/msg14673.html Currently this is a source distribution only; we'll see about putting together some binary distributions after the holidays. ------------------------------ Date: Thu, 20 Dec 2007 07:04:23 GMT From: Malcolm Dunnett Subject: Re: quick question Message-ID: Bob Kaplow wrote: > The M-5 computer or the M-5 multitronic unit was created by Doctor Richard > Daystrom. It utilized very sophisticated technology, probably similar to the > Human neural network, and much more sophisticated than the duotronic > computer commonly in use. According to Dr. Daystrom, the computer could > think and reason like a Human. He had used his own memory engrams as a model > for the computer. Models M-1 through M-4 had been total failures. > Yeah, but it was no match for Bill Shatner!!!!! ------------------------------ Date: 20 Dec 2007 07:30:15 -0600 From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) Subject: Re: quick question Message-ID: In article , Malcolm Dunnett writes: > > Yeah, but it was no match for Bill Shatner!!!!! JTK had a habit of talking computers into commiting suicide. Maybe we should let him loose in Redmond. ------------------------------ Date: Thu, 20 Dec 2007 07:26:03 -0800 (PST) From: AEF Subject: Re: quick question Message-ID: On Dec 20, 9:30 am, koeh...@eisner.nospam.encompasserve.org (Bob Koehler) wrote: > In article , Malcolm Dunnett writes: > > > Yeah, but it was no match for Bill Shatner!!!!! > > JTK had a habit of talking computers into commiting suicide. Maybe > we should let him loose in Redmond. The only other episode I can recall offhand in which this type of thing happens is The Changeling (Nomad). The end of the show where Kirk talks Nomad into self-destruction is absolutely brilliant. What other episodes am I forgetting? There's Landru (Are you of the body? :-) but I think they just used their phasers on it. AEF ------------------------------ Date: Thu, 20 Dec 2007 15:44:00 +0000 (UTC) From: m.kraemer@gsi.de (Michael Kraemer) Subject: Re: quick question Message-ID: In article , AEF writes: > On Dec 20, 9:30 am, koeh...@eisner.nospam.encompasserve.org (Bob > Koehler) wrote: > > In article , Malcolm Dunnett > writes: > > > > > Yeah, but it was no match for Bill Shatner!!!!! > > > > JTK had a habit of talking computers into commiting suicide. Maybe > > we should let him loose in Redmond. > > The only other episode I can recall offhand in which this type of > thing happens is The Changeling (Nomad). The end of the show where > Kirk talks Nomad into self-destruction is absolutely brilliant. What > other episodes am I forgetting? There's Landru (Are you of the > body? :-) but I think they just used their phasers on it. > > AEF ISTR a similar episode where the computer was talked into the task of calculating PI up to the last digit ... It was kind of exorcism because some daemon had taken over the main computer. But I also STR a later episode where the (newer) computer was smart enough to recognize that this is a mission impossible and refused to do it. ------------------------------ Date: Thu, 20 Dec 2007 03:42:28 -0800 (PST) From: Bob Gezelter Subject: Re: Volume label. Message-ID: <4284eefa-d220-49df-aaa3-70f12a92bfa6@i12g2000prf.googlegroups.com> On Dec 19, 2:12 pm, Chuck Aaron wrote: > I have a ds10 with 2 disks. One is named alphasys which is the > system disk and one named userdisk (which equates to user$disk). > Is there a way to add an alias name to user$disk so that when > an application wants to access web$disk it will access the > user$disk? > > Thanks in advance. Chuck, As has been noted, logical names are the answer to this need. For safety's sake, my standard recommendation to clients with a situation such as this is to define a logical name WEB$DISK in terms of the already existent name for USER$DISK making use of DCL Lexical functions, to wit: $ ASSIGN [various qualifiers as appropriate] 'F $TRNLNM("USER$DISK")' WEB$DISK Logical names, and for that matter, DCL lexical functions, are one of the most powerful tools on OpenVMS. A thorough read of the relevant manuals such as the DCL Dictionary, Programming Concepts, and System Services manuals (all of which are available on the OpenVMS WWW site at http://www.hp.com/go/openvms ) is highly recommended. I also published a series of columns on the use of logical names on OpenVMS.org. The first of these "The OpenVMS Consultant: Logical Names (Part 1)", is available at http://www.openvms.org/stories.php?story=02/09/24/5441505 The power of logical names was also the subject of my OpenVMS Technical Journal paper " Inheritance Based Environments in Stand- alone OpenVMS Systems and OpenVMS Clusters", a reprint of which is available from either the OpenVMS WWW site or my site at http://www.rlgsc.com/publications/vmstechjournal/inheritance.html There are also several presentations on related topics available from my various DECUS presentations, available via http://www.rlgsc.com/presentations.html - Bob Gezelter, http://www.rlgsc.com ------------------------------ Date: Thu, 20 Dec 2007 07:34:45 -0800 (PST) From: AEF Subject: Re: Volume label. Message-ID: On Dec 20, 7:42 am, Bob Gezelter wrote: > On Dec 19, 2:12 pm, Chuck Aaron wrote: > > > I have a ds10 with 2 disks. One is named alphasys which is the > > system disk and one named userdisk (which equates to user$disk). > > Is there a way to add an alias name to user$disk so that when > > an application wants to access web$disk it will access the > > user$disk? > > > Thanks in advance. > > Chuck, > > As has been noted, logical names are the answer to this need. For > safety's sake, my standard recommendation to clients with a situation > such as this is to define a logical name WEB$DISK in terms of the > already existent name for USER$DISK making use of DCL Lexical > functions, to wit: > > $ ASSIGN [various qualifiers as appropriate] 'F > $TRNLNM("USER$DISK")' WEB$DISK That depends on his particular needs, no? Some might instead need something like $ DEFINE /SYSTEM/EXEC WEB$DISK USER$DISK: so that reDEFINEing USER$DISK would automatically reDEFINE both logical names. It depends on your what your site needs. Be sure to include the colon in the equivalence name in this command (you should *NOT* use the colon inside the F$TRNLNM function as in Bob Gezelter's example above)! Otherwise you may get problems with certain commands (like SET DEFAULT). You should always include the trailing colon in the equivalence name whenever it is a disk or another logical name that ultimately translates to a generalized file specification. [...] > - Bob Gezelter,http://www.rlgsc.com AEF ------------------------------ Date: Thu, 20 Dec 2007 06:57:27 GMT From: Malcolm Dunnett Subject: Re: Why can't I use an IA64 as a boot server for an Alpha? Message-ID: JF Mezei wrote: > The way it was described, it *appears* that code was added to disable > coss platform booting. > As I said earlier, I suspect that may be true, but not for any nefarious reason. I imagine the idea is that in a mixed architecture cluster one might want to share a common LANCP database, but you wouldn't want the IA64 systems to offer to service the Alpha nodes if you had an Alpha boot server available to do this. I'm sure that at the worst they just never considered anyone would want to do this - and to be fair there are fairly few situations where doing so makes a lot of sense. ------------------------------ Date: Thu, 20 Dec 2007 06:30:58 -0800 (PST) From: etmsreec@yahoo.co.uk Subject: Re: Why can't I use an IA64 as a boot server for an Alpha? Message-ID: <1d5ebda8-0809-4b33-9096-a9cbf8230a36@d21g2000prf.googlegroups.com> On 20 Dec, 06:57, Malcolm Dunnett wrote: > JF Mezei wrote: > > The way it was described, it *appears* that code was added to disable > > coss platform booting. > > As I said earlier, I suspect that may be true, but not for any > nefarious reason. I imagine the idea is that in a mixed architecture > cluster one might want to share a common LANCP database, but you > wouldn't want the IA64 systems to offer to service the Alpha nodes if > you had an Alpha boot server available to do this. > > I'm sure that at the worst they just never considered anyone would > want to do this - and to be fair there are fairly few situations where > doing so makes a lot of sense. Integrity servers use more "industry standard" (whatever they are) protocols to do satellite booting. This is probably so that if DECnet's version/control of MOP or LANCP broke then there wouldn't be the same issues as if Integrity relied on it. Integrity is different because it's not an in-house design from Digital/Compaq so there isn't the same capability to be able to define what it should do, how it should do it, how to fit things in and such like. With disks being "cheap as chips" these days there may not be the same emphasis on booting systems without (in)directly attached disks anyway. Customers may wish to boot from SAN or SmartArray or real disks rather than booting off a boot server across the network. TCP/IP is also the "strategic" network protocol for OpenVMS. It would probably be difficult to get TCP/IP Services to control MOP DLL booting (if for no other reason than Weendoze people take DLL to mean Run Time Library!) Steve ------------------------------ End of INFO-VAX 2007.696 ************************