1 INFO-VAX	Fri, 12 Aug 2005	Volume 2005 : Issue 448       Contents:* Re: $SHOW DEVICE/FILES enhancement request* Re: $SHOW DEVICE/FILES enhancement request* Re: $SHOW DEVICE/FILES enhancement request* Re: $SHOW DEVICE/FILES enhancement request. Re: 7.3-2 Crashes when a Samba Client connects. Re: 7.3-2 Crashes when a Samba Client connects Re: Am I Running OpenSSL?   Favorite Operating System Survey= HP0-436 OpenVMS V7 Network Administration Exam / Study Guides  Image restore fails ( Re: looking for TIMES-ROMAN.XDPS$OUTLINE( Re: looking for TIMES-ROMAN.XDPS$OUTLINE  Re: Memory Reclamation Questions Re: Moving System Disk Re: Moving System Disk strange MOUNT problem  Re: strange MOUNT problem  Re: strange MOUNT problem  Re: strange MOUNT problem 3 TCPIP Services: SMTP bug/wrong logic with 450 codes  trainer for openvms  Re: trainer for openvms  Re: trainer for openvms  Turning up an Ethernet & tftp   F ----------------------------------------------------------------------    Date: 12 Aug 2005 07:25:49 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 3 Subject: Re: $SHOW DEVICE/FILES enhancement request 3 Message-ID: <4Mqwou8dJ+SQ@eisner.encompasserve.org>   c In article <$lM58K3+NdK0@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: s > In article <J3YVqJPJIbki@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes:  > K >>    Me thinks you mean SYSMAN.  Yep, good stuff.  I just never understood A >>    why SYSMAN required OPER privilege, even for such otherwize  >>    non-privileged actions.  > E > Would the person running SYSMAN have been permitted to log into the D > other node in the cluster directly ?  What kind of connections areE > available to that node and are they the sort over which this person  > is allowed to log in ?  B    In these cases the cluster was operating homogeneously (I don'tA    think non-homogeneous clusters are supported anymore).  Anyone D    who could log into any node could log into any other node, submitE    batch jobs to queues on any node, access any node via SCS, DECnet, E    LAT, or IP, or create detached processes on any node using RUN/ON.   F    I do think that requiring OPER when logins are turned off just likeH    interactive logins is appropriate, as well as not providing any other,    exception to existing privilege controls.   ------------------------------  % Date: Fri, 12 Aug 2005 08:49:47 -0400 + From: Ken Robinson <kenrbnsn1@patmedia.net> 3 Subject: Re: $SHOW DEVICE/FILES enhancement request < Message-ID: <6.2.5.2.2.20050812084542.07c72dc8@patmedia.net>  3 At 08:25 AM 8/12/2005, Bob Koehler wrote (in part): D >    In these cases the cluster was operating homogeneously (I don't; >    think non-homogeneous clusters are supported anymore).   F Sure they are. You can have a cluster where each machine has it's own D system disk. At my last position, we had a cluster of 4 machines, 3 F shared a system disk, 1 had it's own. All were running v7.3-2, but we D installed the latest patches on the node with it's own system drive E first, then installed them on to the other drive after a week or two.    Ken Robinson   ------------------------------    Date: 12 Aug 2005 09:22:06 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) 3 Subject: Re: $SHOW DEVICE/FILES enhancement request 3 Message-ID: <jIXxN3ZAwI9g@eisner.encompasserve.org>   q In article <4Mqwou8dJ+SQ@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: e > In article <$lM58K3+NdK0@eisner.encompasserve.org>, Kilgallen@SpamCop.net (Larry Kilgallen) writes: t >> In article <J3YVqJPJIbki@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: >>  L >>>    Me thinks you mean SYSMAN.  Yep, good stuff.  I just never understoodB >>>    why SYSMAN required OPER privilege, even for such otherwize >>>    non-privileged actions. >>  F >> Would the person running SYSMAN have been permitted to log into theE >> other node in the cluster directly ?  What kind of connections are F >> available to that node and are they the sort over which this person >> is allowed to log in ?  > D >    In these cases the cluster was operating homogeneously (I don't; >    think non-homogeneous clusters are supported anymore).   E The MicroVAX II is not supported any more, but it has been repeatedly C explained that such statements do _not_ mean code exclusive to that E system has been removed, just that it is no longer in the text matrix ? and they do not accept problem reports specific to that system.   B It would be a very bad move to change the security behavior of VMSD in response to such a "supportedness" argument.  Remember, the firstD version of VMS evaluated at C2 was VMS V4.3.  The very next release,/ VMS V4.4, introduced a monstrous security hole.    ------------------------------    Date: 12 Aug 2005 12:39:30 -0500 From: briggs@encompasserve.org3 Subject: Re: $SHOW DEVICE/FILES enhancement request 3 Message-ID: <RqSJsK10pYse@eisner.encompasserve.org>   q In article <G16RoA2+TFWn@eisner.encompasserve.org>, koehler@eisner.nospam.encompasserve.org (Bob Koehler) writes: l > In article <6.2.5.2.2.20050812084542.07c72dc8@patmedia.net>, Ken Robinson <kenrbnsn1@patmedia.net> writes:6 >> At 08:25 AM 8/12/2005, Bob Koehler wrote (in part):F >>>    In these cases the cluster was operating homogeneously (I don't= >>>    think non-homogeneous clusters are supported anymore).  >>  I >> Sure they are. You can have a cluster where each machine has it's own  G >> system disk. At my last position, we had a cluster of 4 machines, 3  I >> shared a system disk, 1 had it's own. All were running v7.3-2, but we  G >> installed the latest patches on the node with it's own system drive  H >> first, then installed them on to the other drive after a week or two. > G >    We had lots of system disks.  IIRC homgeneous clustering means one B >    SYSUAF, one RIGHTSLIST, ..., no matter how many system disks.  B Matches my recollection.  DEC is not obliged to help you deal withE BATCH queue issues if you submit a job on a node using one SYSUAF and A it actually runs on a node with a different SYSUAF.  Nor are they ? obliged to help if you put an ACL on a file on a disk using one < RIGHTSLIST and then wonder why it doesn't control access the: way you expected on another node using another RIGHTSLIST.  
 "Unsupported"   B Some of us purists would prefer to use the terms "homogeneous" andI "heterogeneous" in a broader way more as end points on a many-dimensional = spectrum than as specifically definable ways of configuring a E particular cluster.  It's certainly clear that there is more than one  way to understand the terms.   	John Briggs   ------------------------------  % Date: Fri, 12 Aug 2005 09:00:41 -0400 , From: "Richard Whalen" <WhalenR@process.com>7 Subject: Re: 7.3-2 Crashes when a Samba Client connects * Message-ID: <ddi6i2$v9$1@news.process.com>  @ The following commands will configure SMBD properly on MultiNet:   $ multinet config/servers  Add smbd tcp  139   samba_root:[bin]smbd_Startup.com select smbd  set flags ucx_Server write  exit  A "Marty Kuhrt" <kuhrt@nospammy.encompasserve.org> wrote in message - news:cP6xMSc8BFZp@eisner.encompasserve.org... E > In article <96ADA4779wspenceraporg@216.168.3.30>, noone@nowhere.com  (Warren Spencer) writes:H > > We've got a fix now (tell mulinet to use the ucx driver).  Thanks to those  > > who pondered this! > >  > > ws > B > Can you document what you did to make it work?  I was just aboutF > to try and start a Samba install on a similar setup (V7.3-2, Mu4.4A,- > SMB2.2.8) and am now more nervous about it.  >  > > 8 > > wspencer@ap.dontspamme.org (Warren Spencer) wrote in* > > <96ACA7719wspenceraporg@216.168.3.30>: > > 
 > >>Hi Folks,  > >>L > >>I've got an OpenVMS crash I need to sort out and I'm looking for help on > >>where to go next.  > >>H > >>This is an OpenVMS 7.3-2 box, with all (applicable) patches applied.J > >>The tcpip stack is Multinet 4.4A, with all applicable patches applied. > >>The problem is this: > >>L > >>When any Samba client attempts to connect to this box, Multinet fires upH > >>smbd_startup.com (as it should) and the box crashs with "SSRVEXCEPT,) > >>Unexpected system service exception".  > >>I > >>The Samba version is 2.2.8, freshly compiled & installed on this box.  > >> > >>Some other info: > >>K > >>- Either a local or remote samba client connecting will cause the crash : > >>- The crash always happens on the first client connect@ > >>- The crash always happens - Samba has yet to work correctlyI > >>- I've configured smbd as a Multinet service, so Multinet is invoking K > >>smbd_startup.com.  But I'm not sure if this is the way it's supposed to  > >>be set up. > >>1 > >>Any ideas where to go next?  What to look at?  > >> > >>Many thanks! > >> > >>ws > >> > >    ------------------------------  % Date: Fri, 12 Aug 2005 11:38:50 -0400 - From: "John E. Malmberg" <wb8tyw@qsl.network> 7 Subject: Re: 7.3-2 Crashes when a Samba Client connects 1 Message-ID: <4tOdnTsf9poRX2HfRVn-2A@adelphia.com>    Richard Whalen wrote: B > The following commands will configure SMBD properly on MultiNet: >  $ multinet config/servers  Add smbd tcp  139   samba_root:[bin]smbd_Startup.com select smbd  set flags ucx_Server write  exit  E Can you post this message on the SAMBA-VMS(at)SAMBA.ORG mailing list  = where Jean-Yvves Collot may see it to put it on his web page?   D There was someone else posting on that list a few months ago with a 7 similar problem and no one posted an answer that I saw.   I Instructions on how to get the SAMBA-VMS mailing list are on a link from  F http://www.samba.org.  If you prefer a news reader, the SAMBA mailing H list is also available on the private news server news.gmane.org as the " gmane.network.samba.vms newsgroup.  D I did not know the answer about this problem, but tried to post the I information about the SAMBA-VMS mailing list/newsgroup in this thread on  D August 8, and apparently something on the ISP side deleted the post % before it got put on the news server.   I Also to all SAMBA users, when posting problems about SAMBA 2.2.*,  There  E have been several SAMBA 2.2.8 releases.  The date encoded in the zip  F file name is important to doing most troubleshooting, and if your zip H file name does not have a date encoded in the zip file name, it is very  ancient.   -John  wb8tyw@qsl.network Personal Opinion Only    ------------------------------    Date: 12 Aug 2005 03:16:01 -0700, From: "Cluster-Karl" <karl.rohwedder@gmx.de>" Subject: Re: Am I Running OpenSSL?C Message-ID: <1123841761.280037.105830@g49g2000cwa.googlegroups.com>    Scott,  A 1st check PROD SH PROD *SSL*, there is the HP product SSL and the  OPENSSL product.  F If you can find some SSL$... logicals, the HP SSL product is installed , E if you find some SSL... logicals without the $, Openssl is installed.   
 regards Kalle    ------------------------------    Date: 12 Aug 2005 09:28:28 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen) ) Subject: Favorite Operating System Survey 3 Message-ID: <8Ah406jiE12H@eisner.encompasserve.org>   ' http://www.thinkgeek.com/feedback.shtml    ------------------------------    Date: 12 Aug 2005 07:44:00 -07005 From: "robert_kersey@bat.com" <robert_kersey@bat.com> F Subject: HP0-436 OpenVMS V7 Network Administration Exam / Study GuidesC Message-ID: <1123857840.853888.205300@g44g2000cwa.googlegroups.com>   @ I am looking to take the HP0-436 Exam in the next month and have started studying for this exam.   B In the EPG for this exam on page 8 there is some recommended study guides / online labs.   E This link does not seem to work. I have had a look in the HP learning . centre and can not find anything there either.   Do any of you have the guides?   The 2 guides are:   - TCP/IP Services for OpenVMS 7.3 Student Guide , TCP/IP Services for Compaq OpenVMS Lab Guide     Thanks in advance    Rob    ------------------------------  % Date: Fri, 12 Aug 2005 12:40:12 -0400 0 From: Peter Sjoberg <peters38@techwiz.spamno.ca> Subject: Image restore fails> Message-ID: <pan.2005.08.12.16.40.11.428997@techwiz.spamno.ca>  H I have  a not so critical system that lost the system disk. When it came> to restoring it I it a problem, the image restored is corrupt.G After testing a lot I think the problem is because of the way I did the   backup with different /Blocksize  @ The way it's done is that first some db files are backed up with something like  9 BACKUP/EXCLUDE=(some,files,*tmp)/GROUP=100/NOCRC/RECORD -    DBDISK:[DATA.DB]*.* - /   BACKTAPE:C1050001.CKP/TAPE_EXPIR=18-MAY-2005    2 Later on the same tape a image backup is done with7 BACKUP/NOCRC/BLOCKS=32767/GROUP=100/IGNORE=(INTERLOCK)- /  /RECORD/JOURNAL=BACKUP_JOURNAL:DEVSYS_040518 -   DEVSYS:/IMAGE -  _$50$MKC500:DEVSYS_040518 -  /TAPE_EXPIR=18-MAY-2005    & Now when I restore the second one with)  backup mkd500:devsys_040518. dkd0:/Image J it looks so fine (no errors) but when I mount it lots of stuff is missing.( Ana/disk gives a long list of stuff likeE %ANALDISK-W-BADDIRENT, invalid file identification in directory entry          [SYS0]SYSCOMMON.DIR;1 = -ANALDISK-I-BAD_DIRHEADER, no valid file header for directory E %ANALDISK-W-BADDIRENT, invalid file identification in directory entry           [000000]VMS$COMMON.DIR;1= -ANALDISK-I-BAD_DIRHEADER, no valid file header for directory 2 %ANALDISK-W-LOSTHEADER, file (16,1,0) SYSEXE.DIR;1          not found in a directory   and of course nothing works.I I did try to mount the tape with /block=32767/record=32767 but no change.   + Any suggestions on how to restore the disk?    /ps    ------------------------------  + Date: Fri, 12 Aug 2005 07:51:06 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)1 Subject: Re: looking for TIMES-ROMAN.XDPS$OUTLINE $ Message-ID: <ddhkda$jcj$1@online.de>  C In article <1123812303.332894.253350@z14g2000cwz.googlegroups.com>, ( "Erik" <erik_g@abri.une.edu.au> writes:   6 > I had the same problem after upgrading to VMS 7.3.1.* > My solution was to copy all the files in- > SYS$COMMON:[SYSFONT.XDPS.OUTLINE]  on a 7.2 " > system to the VMS 7.3.1 system.   2 OK.  Can someone tell me where to get these fonts?  F I understand the problem with the licensing for Display PostScript andF why it was removed (unfortunate, at least for those who use(d) it, butH apparently a decision made at Adobe).  However, if I just need the fontsC (or, rather, GhostScript needs the fonts), then is there any reason G (legal or technical) why I can't have them?  Assuming that they should  H be unavailable; is there any way to get GhostScript do what I wanted it ) to do (convert a PostScript file to PDF)?    ------------------------------  % Date: Fri, 12 Aug 2005 10:11:01 -0700 # From: "Tom Linden" <tom@kednos.com> 1 Subject: Re: looking for TIMES-ROMAN.XDPS$OUTLINE ( Message-ID: <opsvencnwezgicya@hyrrokkin>  K On Fri, 12 Aug 2005 07:51:06 +0000 (UTC), Phillip Helbig---remove CLOTHES   1 to reply <helbig@astro.multiCLOTHESvax.de> wrote:   E > In article <1123812303.332894.253350@z14g2000cwz.googlegroups.com>, ) > "Erik" <erik_g@abri.une.edu.au> writes:  > 7 >> I had the same problem after upgrading to VMS 7.3.1. + >> My solution was to copy all the files in . >> SYS$COMMON:[SYSFONT.XDPS.OUTLINE]  on a 7.2" >> system to the VMS 7.3.1 system. > 4 > OK.  Can someone tell me where to get these fonts? > H > I understand the problem with the licensing for Display PostScript andH > why it was removed (unfortunate, at least for those who use(d) it, butJ > apparently a decision made at Adobe).  However, if I just need the fontsE > (or, rather, GhostScript needs the fonts), then is there any reason H > (legal or technical) why I can't have them?  Assuming that they shouldI > be unavailable; is there any way to get GhostScript do what I wanted it + > to do (convert a PostScript file to PDF)?  > F anonympous ftp to my site under dir pub.misc you will find outline.zip   ------------------------------  % Date: Fri, 12 Aug 2005 05:34:53 -0700 ( From: Jeff Cameron <roktsci@comcast.net>) Subject: Re: Memory Reclamation Questions 0 Message-ID: <BF21E57D.12347%roktsci@comcast.net>   On 8/11/05 7:30 PM, in article; 1123813845.852112.86090@z14g2000cwz.googlegroups.com, "AEF"  <spamsink2001@yahoo.com> wrote:    > I > Just some curiosity questions. I'm not trying to solve some performance 
 > problem. > B > With proactive memory reclamation turned on, periodically wakingH > processes are trimmed by 25% if they are inactive 99% of the time in aG > 30 sec. interval. OK. What is the limit for how small the working set H > be reduced by repeated trimmings (assuming the free list remains belowG > 2*FREEGOAL)? I've seen values well below SWPOUTPGCNT (500) on some of D > my systems. Is the final limit MINWSCNT or AWSMIN? Something else? > E > Also, with proactive memory reclamation, long waiting processes are G > trimmed and/or swapped out. For both this and the periodically waking G > processes, pages are removed from the working set but the working set I > limit is not reduced. Now my question is, is the working set limit also G > left alone for regular old-fashioned swapper trimming, the kind where F > memory drops below FREELIM? IOW: If the following hold: memory dropsI > below FREELIM, proactive mem. reclamation is turned on -- are processes I > trimmed but their working set LIMITS left alone, or are they left alone I > only for trimming when the free list is less than 2*FREEGOAL but is not * > recovering from being less than FREELIM? >  > 	 > Thanks.  > I Unless you are really hurting for memory, like less than 64 megs, I would J not use this feature. All it does is trim memory that is not needed makingH your system less responsive. I once had an admin who thought this was soH important to trim memory working sets, he didn't realize that he had allF sorts of free memory. I turned off the feature, and all the users were1 saying the system was many times more responsive.   
 Just my $0.02  Jeff   ------------------------------  % Date: Fri, 12 Aug 2005 07:37:06 -0500 + From: brandon@dalsemi.com (BRANDON, JOHN M)  Subject: Re: Moving System Disk 1 Message-ID: <05081207370692@dscis6-0.dalsemi.com>   $ 1) Put both system disks on the SAN.  G Faster, and more reliable if the SAN and disks are configured properly. ; I have had problems with internal disks with software RAID.   K If you were on the SAN right now you would be able to swap the servers with O little change requirements.  Simply adjust the SAN connections and WWID on your  server.   N There are other advantages - being able to work on the disk (if another serverK is available) if you somehow manage to trash it or you screw up the startup  process.  ( ESA, EMA, EVA series all work with this.  F 2) There are hardware specific versions when new hardware is released.  8 However that release is rolled up into the next version.@ For example, V7.2-1h1 for the GS series was rolled up into V7.3.A You have no concern since you are running V7.3-2 on both systems.   0 3) Image backup from disk to new-disk will work.  M An AUTOGEN would be recommended (if not required) to adjust some parameters.  & Especially if more memory is involved.     John "REBOOT" Brandon  VMS Systems Administrator * firstname.lastname.spam.me.not@dalsemi.com   ------------------------------  % Date: Fri, 12 Aug 2005 09:13:15 -0400 * From: "Syltrem" <syltremzulu@videotron.ca> Subject: Re: Moving System Disk 0 Message-ID: <2g1Le.121$2Q.218@tor-nn1.netcom.ca>  > "Dave Froble" <davef@tsoft-inc.com> a crit dans le message de* news:11fnhuro2gir2c0@corp.supernews.com... > Ed Wilts wrote: D > >>We are moving from an ES40 with a local system disk and many SANI > >>attached disks to a GS80 with a local system disk that will attach to K > >>the same SAN disks the ES40 currently attaches to, after the migration. K > >> Both are running VMS 7.3-2.  What is the preferred method of migrating & > >>the system disk to the new server? > >  > > J > > The first question is why not boot off the SAN?  It works like a charmH > > - I just finished migrating my production cluster from CI storage toE > > SAN storage and it's not as hard as you might think if you have a > > > volume shadowing license (or can get a temporary license). > D > What advantages do you see in putting your system disk on the SAN? >     D One advantage is that you can mirror it without a Shadowing license.H Another is that disks on the SAN are a lot (a lot!) faster than the SCSI connected drives. F At least this is true for my ES40 here. I only use the local disks for1 storage (not for files heavily accessed by users) K And at last, you can boot many nodes from the same system disk, if you have * a cluster and the system disk is on a SAN.   --   Syltrem    OpenVMS 7.3-1 + Oracle 9.2.0.5H http://pages.infinit.net/syltrem (OpenVMS related web site, en franais)% ---zulu is not in my email address---     G > I'd want the capability of bringing up my VMS system without the SAN, 1 > should the SAN not be available for any reason.  > H > What good would it be without the applications?  Don't know.  I'd just > want the capability. > I > As you mentioned, a local disk can also be set up as a alternate system  > disk.  >  > --  6 > David Froble                       Tel: 724-529-04506 > Dave Froble Enterprises, Inc.      Fax: 724-529-0596@ > DFE Ultralights, Inc.              E-Mail: davef@tsoft-inc.com > 170 Grimplin Road  > Vanderbilt, PA  15486    ------------------------------  + Date: Fri, 12 Aug 2005 07:44:20 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: strange MOUNT problem$ Message-ID: <ddhk0k$j6b$1@online.de>   $ sh dev di   P Device                  Device           Error    Volume         Free  Trans MntP  Name                   Status           Count     Label        Blocks Count CntP $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3                         mounted + $ mount/cluster/noassist $5$DIA3: SCRATCH_2 9 %MOUNT-I-MOUNTED, SCRATCH_2 mounted on _$5$DIA3: (FLAMME)  $ sh dev di   P Device                  Device           Error    Volume         Free  Trans MntP  Name                   Status           Count     Label        Blocks Count CntP $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3                         mounted   E The node from which the command is issued is the node which hosts the D disk (actually, it is a DSSI bus, but there is only one node in the H cluster with DSSI and that is on the node from which the above commands 
 were issued).    SYSMAN> do sh dev di2 %SYSMAN-I-OUTPUT, command execution on node GLADIAP Device                  Device           Error    Volume         Free  Trans MntP  Name                   Status           Count     Label        Blocks Count CntP $5$DIA3:      (DANEEL)  MntVerifyTimeout     0  SCRATCH_2      1177413     1   32 %SYSMAN-I-OUTPUT, command execution on node DANEELP Device                  Device           Error    Volume         Free  Trans MntP  Name                   Status           Count     Label        Blocks Count CntP $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3                         mounted 2 %SYSMAN-I-OUTPUT, command execution on node ELIJAHP Device                  Device           Error    Volume         Free  Trans MntP  Name                   Status           Count     Label        Blocks Count CntP $5$DIA3:      (DANEEL)  Mounted              0  SCRATCH_2      1177413     1   3  E I have no idea why the disk should be in MntVerifyTimeout.  Yesterday G everything was fine and the last reboot was a few days ago (and on the  < node where there is MntVerifyTimeout it's been even longer).  0 SYSMAN> do write sys$output f$getsyi("boottime")2 %SYSMAN-I-OUTPUT, command execution on node GLADIA 23-JUL-2005 08:07:04.00 2 %SYSMAN-I-OUTPUT, command execution on node DANEEL  6-AUG-2005 08:04:09.48 2 %SYSMAN-I-OUTPUT, command execution on node ELIJAH  7-AUG-2005 08:46:41.95   D What really puzzles me is that it is HostUnavailable on the node to F which it is connected (via the DSSI bus).  How can it be mounted on a D node to which it is just MSCP served but unavailable on the node to 6 which it has a "direct" connection (via the DSSI bus)?  H From the node where it shows up as mounted, I tried a DIR command.  The # disk is now in MountVerify there.     ? Is there any way to get out of this situation without a reboot?   B I'm not on-site at the moment and usually reboot via the physical G console.  I don't have any remote consoles.  Can I log in via LAT from  C node A to node B and do @SYS$SYSTEM:SHUTDOWN on node B, specifying  D REBOOT, and expect it to work?  Or do I need to do the shutdown and B reboot via SYSMAN (and will this work).  Obviously, when the node B reboots, it kills all processes, including the one from which the C shutdown was specified.  Obviously, from the console this is not a  H problem, but I'm wondering if a LAT connection would be OK (there is no F DECnet running; TCPIP is running if that wouold be preferable to LAT).   MVTIMEOUT is 3600 on all nodes.    ------------------------------    Date: 12 Aug 2005 03:11:59 -0700, From: "Cluster-Karl" <karl.rohwedder@gmx.de>" Subject: Re: strange MOUNT problemB Message-ID: <1123841519.728261.72850@g43g2000cwa.googlegroups.com>  1 Phillip Helbig---remove CLOTHES to reply schrieb:   
 > $ sh dev di  > R > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count CntR > $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3! >                         mounted - > $ mount/cluster/noassist $5$DIA3: SCRATCH_2 ; > %MOUNT-I-MOUNTED, SCRATCH_2 mounted on _$5$DIA3: (FLAMME) 
 > $ sh dev di  > R > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count CntR > $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3! >                         mounted  > G > The node from which the command is issued is the node which hosts the E > disk (actually, it is a DSSI bus, but there is only one node in the I > cluster with DSSI and that is on the node from which the above commands  > were issued).  >  > SYSMAN> do sh dev di4 > %SYSMAN-I-OUTPUT, command execution on node GLADIAR > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count CntR > $5$DIA3:      (DANEEL)  MntVerifyTimeout     0  SCRATCH_2      1177413     1   34 > %SYSMAN-I-OUTPUT, command execution on node DANEELR > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count CntR > $5$DIA3:      (FLAMME)  HostUnavailable      0  SCRATCH_2      1177413     1   3! >                         mounted 4 > %SYSMAN-I-OUTPUT, command execution on node ELIJAHR > Device                  Device           Error    Volume         Free  Trans MntR >  Name                   Status           Count     Label        Blocks Count CntR > $5$DIA3:      (DANEEL)  Mounted              0  SCRATCH_2      1177413     1   3 > G > I have no idea why the disk should be in MntVerifyTimeout.  Yesterday H > everything was fine and the last reboot was a few days ago (and on the> > node where there is MntVerifyTimeout it's been even longer). > 2 > SYSMAN> do write sys$output f$getsyi("boottime")4 > %SYSMAN-I-OUTPUT, command execution on node GLADIA > 23-JUL-2005 08:07:04.00 4 > %SYSMAN-I-OUTPUT, command execution on node DANEEL >  6-AUG-2005 08:04:09.48 4 > %SYSMAN-I-OUTPUT, command execution on node ELIJAH >  7-AUG-2005 08:46:41.95  > E > What really puzzles me is that it is HostUnavailable on the node to G > which it is connected (via the DSSI bus).  How can it be mounted on a E > node to which it is just MSCP served but unavailable on the node to 8 > which it has a "direct" connection (via the DSSI bus)? > I > From the node where it shows up as mounted, I tried a DIR command.  The # > disk is now in MountVerify there.  > A > Is there any way to get out of this situation without a reboot?  > C > I'm not on-site at the moment and usually reboot via the physical H > console.  I don't have any remote consoles.  Can I log in via LAT fromD > node A to node B and do @SYS$SYSTEM:SHUTDOWN on node B, specifyingE > REBOOT, and expect it to work?  Or do I need to do the shutdown and C > reboot via SYSMAN (and will this work).  Obviously, when the node C > reboots, it kills all processes, including the one from which the D > shutdown was specified.  Obviously, from the console this is not aI > problem, but I'm wondering if a LAT connection would be OK (there is no H > DECnet running; TCPIP is running if that wouold be preferable to LAT). > ! > MVTIMEOUT is 3600 on all nodes.   G There seems to by a problem with the disk, can you SET HOST/DSSI to it? E (I'm not quite sure about the syntax, DSSI disks are long gone here).   ! A reboot with SYSMAN should work.   
 regards Kalle    ------------------------------  + Date: Fri, 12 Aug 2005 14:37:43 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) " Subject: Re: strange MOUNT problem( Message-ID: <ddic7n$ju8$1@pcls4.std.com>  R helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes:  E >What really puzzles me is that it is HostUnavailable on the node to  G >which it is connected (via the DSSI bus).  How can it be mounted on a  E >node to which it is just MSCP served but unavailable on the node to  7 >which it has a "direct" connection (via the DSSI bus)?   I >From the node where it shows up as mounted, I tried a DIR command.  The  $ >disk is now in MountVerify there.    F The way DSSI and MSCP Served drives work this is expected.  Each driveG is served by a disk server.  For MSCP Server the disk server is another F VMS host.  For DSSI it's a little disk server on the drive itself or aK "brick" for HSDs.  The SCS nodename of the server is what's in parenthesis.   H If the disk server becomes unavailable you'll get the "host unavailable" message from $ show device.   H Related but separate is "mount verify".  When an IO to a disk fails withH certain conditions, mount verify will try to bring the disk back online,F and it (really the driver) tries other paths to the drive if possible F (other disk servers).  If it succeeds it will verify the drive broughtC online is really the same one as before.  Note this only happens if G an IO fails.  It retries for MVTIMEOUT seconds, in which case the drive 2 goes into MntVerifyTimeout state until dismounted.  F Client DANEEL sees DSSI disk server FLAMME offline, but because it hasG not (apparently) done any I/O, Mount Verify has not been invoked so the B drive remains MOUNTed.  Do any I/O and you'll see it go into mountG verify, at least for a moment.  (I/Os done as proxy for the MSCP server 2 on behalf of other nodes don't count in the above)  G Client ELIJAH sees the disk server DANEEL online, and since it has not  J (apparently) done any I/O, Mount Verify has not been invoked there either.* Things will look happy until it tries I/O.  C Client GLADIA sees the disk server DANEEL online, and it apparently C _has_ tried to do I/O, MV couldn't bring the disk online (via proxy I through DANEEL) and it gave up.  (failed I/Os on the MSCP server get sent G back to the client who does the MV stuff, the server pretty much passes ) along all requests and sends status back)   @ >Is there any way to get out of this situation without a reboot?  8 Try $ DISMOUNT/ABORT on all nodes which have it mounted.   ------------------------------  + Date: Fri, 12 Aug 2005 16:53:54 +0000 (UTC) 7 From: moroney@world.std.spaamtrap.com (Michael Moroney) " Subject: Re: strange MOUNT problem( Message-ID: <ddik72$itp$1@pcls4.std.com>  R helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes:  C >> >Is there any way to get out of this situation without a reboot?  >>  ; >> Try $ DISMOUNT/ABORT on all nodes which have it mounted.   H >I understand what is going on now.  It appears that there is a problem G >with the DSSI server.  Without a reboot of DANEEL (DANEEL is the only  I >node in this cluster with DSSI), is there any way (via SET HOST/DUP) to  H >restart the DSSI server or at least try to connect to it?  What is the > >magic incantation?  Do I have to load any drivers via SYSGEN?  J I don't remember the SET HOST/DUP syntax.  Check the cables between DANEELD and the DSSI disk server.  Reboot the disk server (possibly by powerH cycling it if necessary) if nothing else works.  Once the disk server isH up and able to talk to DANEEL, normal mount or mount verify should bring the drive back.    ------------------------------  % Date: Fri, 12 Aug 2005 03:30:38 -0400 - From: JF Mezei <jfmezei.spamnot@teksavvy.com> < Subject: TCPIP Services: SMTP bug/wrong logic with 450 codes, Message-ID: <42FC5005.68C16F8D@teksavvy.com>  H The VMS SMTP server fails to handle greylisting by the receiving server.A (VAX TCPIP Services 5.3-2). And it gives the used erroneous error N message, and logs don't show the real error message unless you enable tracing.      # The standard (RFC 2821) stipulates:   9 "450 Requested mail action not taken: mailbox unavailable           (e.g., mailbox busy)"  ; More specifically for 400 seried messages: (page 42 of RFC)   + "4yz   Transient Negative Completion reply"   / Which means: this isn't fatal, try again later.     G 450  is used by smart SMTP servers with greylisting implemented to tell E the sending SMTP server to retry in 30 minutes of more. (spam senders  don't bother retrying).   Y =========================================================================================     2 (note:  domain names munged to prevent harverters)  D Part 1 : The trace: (TCPIP$SMTP_SYMB_TRACE set to 1). Note the error$ message given by the BA SMTP server.    D %%%%%%%%%%%%                   12-AUG-2005 02:48:36.66  %%%%%%%%%%%%  G %TCPIP-I-SMTP_SYMBRUN, symbiont is running the queue TCPIP$SMTP_BIKE_01   P 12-AUG-2005 02:49:11.10: next open file _$2$DKA200:[JFMEZEI.MAILBOX]050812024909$ 48_JFMEZEI-20400104.TCPIP_BIKE;1(72)3 recv buf=220 cr02ae01.baplc.com ESMTP Postfix\0d\0a & send buf=HELO bike.vaxination.Ka\0d\0a% recv buf=250 cr02ae01.baplc.com\0d\0a 0 send buf=MAIL FROM:<jfmezei@vaxination.Ka>\0d\0a recv buf=250 Ok\0d\0a 5 send buf=RCPT TO:<WEBMASTER@britishairwayZ.com>\0d\0a > recv buf=450 <jfmezei@vaxination.Ka>: Sender address rejected:: unverified address: Address verification in progress\0d\0a send buf=QUIT\0d\0a  recv buf=221 Bye\0d\0a    [ ===========================================================================================   H Part 2: The normal log file portion: (note the "requeued" even though itD is in fact permanently bounced and not requeued).  Also note the 450F error code in the log file which doesn't match what the BA server said@ and misleads one into thinking britishairways.com doesn't exist.  / Printing debug_level 2, Domains and recipients:  Domain: britishairways.com2    Recipient address: WEBMASTER@BRITISHAIRWAYZ.COM(    Domain part:       britishairwayZ.com    Local part:        WEBMASTER 9    Address Status:    Not done, not delivered. (Requeued)   ) Printing debug_level 2, relevant headers: " Return-Path: jfmezei@vaxination.Ka% Date: Fri, 12 Aug 2005 02:49:09 -0400 * Message-Id: <05081202490948@vaxination.Ka> From: jfmezei@vaxination.Ka   To: WEBMASTER@britishairwayZ.com# Subject: Javascript page not found. & X-VMS-To: webmaster@britishairwayZ.com  @ A permanent error makes this mail undeliverable. Must bounce it.? 450  %TCPIP-E-SMTP_NOSUCHUSER, no such user, britishairways.com   / Printing debug_level 2, Domains and recipients:  Domain: britishairways.com2    Recipient address: WEBMASTER@BRITISHAIRWAYZ.COM(    Domain part:       britishairwayZ.com    Local part:        WEBMASTER 9    Address Status:    Not done, not delivered. (Requeued)   Y =========================================================================================   F Part 3: This is the start of the message the VMS SMTP postmaster sendsF back to the sender: (again, note that it gives the wrong error message9 and points to britishairways.com instead of the username.      % Date: Fri, 12 Aug 2005 02:42:26 -0400 * Message-Id: <05081202422627@vaxination.Ka> From: postmaster@vaxination.Ka To: jfmezei@vaxination.Ka  Subject: Returned mail    ' ---- Transcript of session follows ----   ? 450  %TCPIP-E-SMTP_NOSUCHUSER, no such user, britishairwayZ.com   ---- Unsent message follows ----   etc     E Damm, all this just to report to the BA website webmaster that one of H his junk javascript files is not available which causes their web serverE to redirect to the home page (which requests that unavailable js file D and loops... at a time when BA tells its customers to access its webC site to find out about the cancellation of all its flights at LHR).     E But at least now I have the proof that the VMS SMTP server misbehaves A and doesn't requeue messages that should be requeued (400 seriies J messages) and doesn't give the right error information back to the sender.   ------------------------------    Date: 11 Aug 2005 22:59:33 -0700# From: "elsa" <Elsa.Leena@gmail.com>  Subject: trainer for openvmsC Message-ID: <1123826373.802343.242250@g49g2000cwa.googlegroups.com>    Hi Group Memebers.  F I am Elsa from Chennai, India. I am looking for a professional who canD impart training program on OpenVMS for one of my clients. If you areC interested, please let me know. Reach me at leena@triadinfotech.com ) Is there anyone from India in this group? ! If so, do get in touch with me...    ------------------------------    Date: 12 Aug 2005 06:10:13 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)   Subject: Re: trainer for openvms3 Message-ID: <J+amIUjmulj8@eisner.encompasserve.org>   i In article <1123826373.802343.242250@g49g2000cwa.googlegroups.com>, "elsa" <Elsa.Leena@gmail.com> writes:   H > I am Elsa from Chennai, India. I am looking for a professional who canF > impart training program on OpenVMS for one of my clients. If you areE > interested, please let me know. Reach me at leena@triadinfotech.com   # You might consider also posting at:   + 	http://www.openvms.org/phorum/list.php?f=2   + > Is there anyone from India in this group?   " People from India read that forum.   ------------------------------  # Date: Fri, 12 Aug 2005 13:50:15 GMT 6 From: "Kenneth Farmer" <kfarmer@NOSPAM.spyderbyte.com>  Subject: Re: trainer for openvms@ Message-ID: <rO1Le.118610$Kp2.10328910@twister.southeast.rr.com>  ; "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message  - news:J+amIUjmulj8@eisner.encompasserve.org... M > In article <1123826373.802343.242250@g49g2000cwa.googlegroups.com>, "elsa"    > <Elsa.Leena@gmail.com> writes: > I >> I am Elsa from Chennai, India. I am looking for a professional who can G >> impart training program on OpenVMS for one of my clients. If you are F >> interested, please let me know. Reach me at leena@triadinfotech.com > % > You might consider also posting at:  > , > http://www.openvms.org/phorum/list.php?f=2 > , >> Is there anyone from India in this group?    $ > People from India read that forum.     That's an understatement!  :)   " Thanks for letting her know Larry.       Ken   % _____________________________________ " Kenneth R. Farmer <>< 336-736-73763 www.OpenVMS.org | dba.OpenVMS.org | dcl.OpenVMS.org  HP OpenVMS News and Info     ------------------------------  % Date: Fri, 12 Aug 2005 11:47:50 -0500 % From: Neil Cherry <njc@wolfgang.uucp> & Subject: Turning up an Ethernet & tftp. Message-ID: <slrndfpkll.87m.njc@wolfgang.uucp>  @ I have a nice Alpha sitting around not doing much (it's used forF DECNet support). It has 2 nice 100M interfaces and I just found a rackB mount scsi drive setup with 2 9.1G drives. What I want to do is toF turn this into our lab's tftpd server (security is not an issue in the? lab). The problem is that one of my 100M interfaces won't start F up. Where do I start? Then how do I add the drives to VMS? Pointers to; links (with clues as to what to start with would be greatly  appreciated).    --  C Linux Home Automation         Neil Cherry       ncherry@comcast.net ; http://home.comcast.net/~ncherry/               (Text only) 8 http://hcs.sourceforge.net/                     (HCS II): http://linuxha.blogspot.com/                    My HA Blog   ------------------------------   End of INFO-VAX 2005.448 ************************