1 INFO-VAX	Sun, 06 Aug 2006	Volume 2006 : Issue 434       Contents:5 Re: Authentication for outgoing TCP/IP Services mail?  Re: Speaking of Clusters:  Re: WASD Problems  Re: WASD Problems   F ----------------------------------------------------------------------  % Date: Sun, 06 Aug 2006 01:04:27 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> > Subject: Re: Authentication for outgoing TCP/IP Services mail?, Message-ID: <44D5784C.E24C2894@teksavvy.com>   Paul Sture wrote:  > J > Some time ago, my ISP changed their mail system so that I need to supply# > authentication for outgoing mail    0 > Does similar exist in TCP/IP Services for VMS?   *NO*.     F The VMS SMTP server was designed as a "real" SMTP server that deliversH to many destinations.  In such a context, authentication is meaningless.G  But in a context where you route all mail to a gateway server (such as > your ISPs), then authentication can become realistic (a singleE user/password needed to send all mail via such server). But the TCPIP , Services server is not setup to handle this.  D And usually, an ISP would allow computers using its own IP blocks to, access its SMTP server without restrictions.  E The question of authentication has been asked a number of times here. G And assuming there are people left at TCPIP Services engineering, it is B doubtful they would be given priority on this feature since paying customers don't request it.   C Your best bet is to find an ISP that doesn't require authentication < and/or block port 25 so your own SMTP server can do its job.   ------------------------------  % Date: Sat, 05 Aug 2006 21:18:26 -0400 ( From: Bill Todd <billtodd@metrocast.net>" Subject: Re: Speaking of Clusters:G Message-ID: <25udnVm-YcP_3kjZnZ2dnUVZ_qGdnZ2d@metrocastcablevision.com>    david20@alpha2.mdx.ac.uk wrote: o > In article <1154727563.509509.98940@b28g2000cwb.googlegroups.com>, "geletine" <adaviscg1@hotmail.com> writes: " >> david20@alpha2.mdx.ac.uk wrote:1 >>> The term cluster has many different meanings.  >>> J >>> Beowulf is a HPC (High Performance Computing) cluster ie making use of/ >>> multiple CPUs to run parallel computations. = >>> This is not the same as a HA (High availability) cluster.  >>> G >> in that case linux-ha, HP Serviceguard for Linux,SteelEye LifeKeeper H >> ,GoAhead SelfReliant ,Microsoft Cluster Server, OpenClovis ASP ,HACMPE >> ,Sun Cluster ,Novell Cluster Services for netware, Red Hat Cluster : >> Suite and finally TruCluster i believe are HA clusters. >># >> what type of cluster was ARCnet?  >>	 >> thanks  >> > F > Failover and shared-nothing configurations are good enough for many.Q > The only ones from that list I wopuld class as anywhere near VMS clusters would O > be recent Sun clusters (though I haven't really looked into them and hence am  > going on hearsay) and Tru64.  E Any specific reason you don't include AIX?  While the AIX clustering  D facilities seem somewhat more bolted-on than VMS's do, I'm not sure E there's very much that a VMS cluster can do that at least one of the  B multiple mechanisms on AIX that more or less fall under the HACMP  umbrella can't match.    - bill   ------------------------------  % Date: Sun, 06 Aug 2006 08:54:42 +0930 * From: Mark Daniel <mark.daniel@vsm.com.au> Subject: Re: WASD Problems0 Message-ID: <12dadnedgjmt744@corp.supernews.com>   Dan Williams wrote: A >>This data doesn't make much sense.  Mind you I'm building it on  >>, >>   Compaq C V6.4-008 on OpenVMS Alpha V7.3+ >>   Process Software MultiNet V5.1 Rev A-X  >> >>and  >>, >>   Compaq C V6.5-001 on OpenVMS Alpha V8.2> >>   HP TCP/IP Services for OpenVMS Alpha Version V5.5 - ECO 1 >>G >>with which undoubtably would introduce some differences.  The closest H >>would be the V7.3 and that indicates the PC is in some inprobable partF >>of the logging module which shouldn't be in use in demo mode anyway. >>. >>1) Which version of C do you have installed? >>/., I >>2) Grab the HTTPD914-AXP.ZIP and do a link-build against that to see if G >>it's something in the local build.  Once you've unarchived the object + >>module you can short-circuit the build by  >> >>   $ SET DEFAULT [SRC.HTTPD] >>   $ @BUILD_HTTPD LINK >>   $ @[INSTALL]DEMO  >>J >>3) If you have the HP SSL product installed try a build against that and >>see if the problem remains.  >  > G > I have Compaq C V6.5-001. If I don't build it and just link, it works I > fine. I get the same error if I build with no SSL.  I don't really need E > to re-compile it but it would be nice to know why it's not working.  >  > Thanks > Dan   F Dan and I took this off-line and it didn't take long to establish the I build was failing for a couple of modules when they were being partially  E compiled before aborting with insufficient memory.  This resulted in  I some functions having zero addresses and of course an simple ACCVIO upon  @ activation.  Bump the build account's PGFLQUO up and no problem!  G The build failure was visually detectable at both the compile and link  I stages but not caught by the build procedure so unless you were looking!  E     Interesting that this is the first time in over a decade such an  C issue has been encountered (or at least reported).  Obviously I'll   modify the procedure.    ------------------------------  # Date: Sun, 06 Aug 2006 02:16:23 GMT + From: "Villy Madsen" <Villy.Madsen@shaw.ca>  Subject: Re: WASD Problems- Message-ID: <XhcBg.319531$iF6.61499@pd7tw2no>   M The last comment reminded me that I ran into the same issue when I was first  K setting up WASD on my simulated VAX.  I can't remember what clued me in at  < the time, and it could have been SSH that caused my grief...  : I guess that one can never have PGFLGUO too   large <G> ??  9 I remember when an 8mb 11/750 was a decent sized box.....    sigh   Villy             8 "Mark Daniel" <mark.daniel@vsm.com.au> wrote in message * news:12dadnedgjmt744@corp.supernews.com... > Dan Williams wrote: B >>>This data doesn't make much sense.  Mind you I'm building it on >>> - >>>   Compaq C V6.4-008 on OpenVMS Alpha V7.3 , >>>   Process Software MultiNet V5.1 Rev A-X >>>  >>>and >>> - >>>   Compaq C V6.5-001 on OpenVMS Alpha V8.2 ? >>>   HP TCP/IP Services for OpenVMS Alpha Version V5.5 - ECO 1  >>> H >>>with which undoubtably would introduce some differences.  The closestI >>>would be the V7.3 and that indicates the PC is in some inprobable part G >>>of the logging module which shouldn't be in use in demo mode anyway.  >>> / >>>1) Which version of C do you have installed?  >>>/.,J >>>2) Grab the HTTPD914-AXP.ZIP and do a link-build against that to see ifH >>>it's something in the local build.  Once you've unarchived the object, >>>module you can short-circuit the build by >>>  >>>   $ SET DEFAULT [SRC.HTTPD]  >>>   $ @BUILD_HTTPD LINK  >>>   $ @[INSTALL]DEMO >>> K >>>3) If you have the HP SSL product installed try a build against that and  >>>see if the problem remains. >> >>H >> I have Compaq C V6.5-001. If I don't build it and just link, it worksJ >> fine. I get the same error if I build with no SSL.  I don't really needF >> to re-compile it but it would be nice to know why it's not working. >>	 >> Thanks  >> Dan > H > Dan and I took this off-line and it didn't take long to establish the K > build was failing for a couple of modules when they were being partially  L > compiled before aborting with insufficient memory.  This resulted in some F > functions having zero addresses and of course an simple ACCVIO upon B > activation.  Bump the build account's PGFLQUO up and no problem! > I > The build failure was visually detectable at both the compile and link  K > stages but not caught by the build procedure so unless you were looking!  M > Interesting that this is the first time in over a decade such an issue has  F > been encountered (or at least reported).  Obviously I'll modify the 
 > procedure.     ------------------------------   End of INFO-VAX 2006.434 ************************