INFO-VAX Mon, 08 Jan 2007 Volume 2007 : Issue 15 Contents: Re: Dec VWS Internals ? Re: DS10L disk devices Re: DS10L disk devices Re: DS10L disk devices Re: DS10L disk devices Re: DS10L disk devices Re: SYSMAN suggestion: SET ENV /CLUSTER /EXCLUDE= Re: [ODS2_WIN32] Reading ODS2/ODS5 CDs/volumes on Win32/Win64? ---------------------------------------------------------------------- Date: Sun, 7 Jan 2007 23:50:42 +0000 (UTC) From: moroney@world.std.spaamtrap.com (Michael Moroney) Subject: Re: Dec VWS Internals ? Message-ID: Alan Frisbie writes: >Thanks for these notes, Fred. I have always been a fan of VWS and miss >some of its features that don't exist today. One weird one I remember was a font it had that was 1 pixel wide and 2 or 3 pixels high. With it you could make an itty bitty terminal window. While you couldn't actually read anything when using this font, you could recognize what was going on, just like viewing a terminal screen from a distance. There was some sort of magic with this font (only) where clicking on the terminal window made it switch to a normal font so you could read it, almost like an icon. Some people used terminal windows with that font to monitor OPCOM messages. ------------------------------ Date: 7 Jan 2007 23:45:25 GMT From: Hans Bachner Subject: Re: DS10L disk devices Message-ID: JF Mezei wrote: > Two apprently identical DS10L machines. Only difference is the second > one has 2 IDE drives. > > First one has DQA0 (disk), DQA1, DQB0, DQB1, DVA0 > Second one has DQA0 (disk), DQB0 (disk). > > Is there a way to configure the first machine to have only DQA0 since > that is the only device installed in it and all others are > non-existant devices ? $ MC SYSMAN IO SET EXCLUDE=(DQA1,DQB*) will exclude these devices from autoconfiguration. > Also, DVA0 is shown as an RX26 device. Does this mean that I could > theoretically take the disquette drive from an old VAXstation and put > it in the Alpha ? I don't think so. The VAX floppy drives were SCSI, iirc (and were shown as DK devices starting with some version of VMS I fail to remember...). The DS10L has a PC-type floppy interface. The list of supported options for the DS10L has a floppy/CD-ROM combo drive which works nicely. Hans. ------------------------------ Date: Mon, 08 Jan 2007 00:59:17 -0500 From: JF Mezei Subject: Re: DS10L disk devices Message-ID: <5bd5b$45a1dde1$cef8887a$1631@TEKSAVVY.COM> Hans Bachner wrote: > $ MC SYSMAN IO SET EXCLUDE=(DQA1,DQB*) Wow ! many thannks. I didn't know about that command or some permanent dabase of devices to exclude. Next time I reboot, I'll see if it really works. (the HELP text for it seems to mention this is meant only for certain devices) and doesn't include IDE stuff). >> Also, DVA0 is shown as an RX26 device. BTW, how come this is shows as an available device ? > Disk $10$DVA0: (BIKE), device type RX26, is online, file-oriented device, > shareable, available to cluster, error logging is enabled. > > Error count 0 Operations completed 0 > Owner process "" Owner UIC [SYSTEM] > Owner process ID 00000000 Dev Prot S:RWPL,O:RWPL,G:R,W > Reference count 0 Default buffer size 512 > Total blocks 2880 Sectors per track 18 > Total cylinders 80 Tracks per cylinder 2 > Allocation class 10 Also: $ mount dva0/override=id %MOUNT-F-FORMAT, invalid media format $ Interesting that it doesn't complain about the device not existing at all. Or is there some miniatur3 RX26 hidden under the motherboard that I can't see ? :-) :-) ------------------------------ Date: Sun, 07 Jan 2007 14:19:47 -0500 From: JF Mezei Subject: Re: DS10L disk devices Message-ID: <49afc$45a147fe$cef8887a$21894@TEKSAVVY.COM> Michael Moroney wrote: > It seems the IDE driver creates devices (UCBs) for the possible IDE drives > and then looks to see if there's anything there. If it were up to me, > I'd eliminate the UCB if there's nothing there, or, better yet, never > create it in the first place. Think of SCSI, it doesn't create phantom > drives for every unused SCSI ID on a SCSI bus. Silly me. Someone reminded me via private email that in a cluster not all devices are served. So, from the point of view of BIKE, its own devices had the superfluous devices but CHAIN didn't. So, from the point of view of CHAIN, its own devices had the superfluous devices but BIKE didn't. So the clustering software does not propagate those non-existent empty devices. But it would still be nice if VMS didn't create those devices to begin with. When using F$DEVICE to get a list of disks, once gets a number of nonexistant devices, so you have to test for f$getdvi(x, "AVL") .AND. f$getdvi(x,"SERVED_DEVICE") to isolate those local nonexistant disks. (normally a non available drive would be assumed to be off-node and not that node's responsability) ------------------------------ Date: Sun, 7 Jan 2007 22:51:40 +0000 (UTC) From: moroney@world.std.spaamtrap.com (Michael Moroney) Subject: Re: DS10L disk devices Message-ID: JF Mezei writes: >Michael Moroney wrote: >> It seems the IDE driver creates devices (UCBs) for the possible IDE drives >> and then looks to see if there's anything there. If it were up to me, >> I'd eliminate the UCB if there's nothing there, or, better yet, never >> create it in the first place. Think of SCSI, it doesn't create phantom >> drives for every unused SCSI ID on a SCSI bus. >Silly me. Someone reminded me via private email that in a cluster not all >devices are served. >So, from the point of view of BIKE, its own devices had the superfluous >devices but CHAIN didn't. >So, from the point of view of CHAIN, its own devices had the superfluous >devices but BIKE didn't. >So the clustering software does not propagate those non-existent empty devices. >But it would still be nice if VMS didn't create those devices to begin >with. When using F$DEVICE to get a list of disks, once gets a number of >nonexistant devices, so you have to test for >f$getdvi(x, "AVL") .AND. f$getdvi(x,"SERVED_DEVICE") to isolate those local >nonexistant disks. (normally a non available drive would be assumed to be >off-node and not that node's responsability) Not sure of the need to test whether is served, just see if it's set /NOAVAILABLE (or the device type being "Nonexistent IDE/ATAPI disk") instead. The MSCP server code itself checks if a drive's AVL bit is set and doesn't serve it if not. The DQDRIVER source seems to be on the freeware, maybe I can "fix" that and make the driver available. However, the freeware version is probably old, so doing so may reintroduce previously fixed problems. ------------------------------ Date: Sun, 7 Jan 2007 20:05:39 -0500 From: "William Webb" Subject: Re: DS10L disk devices Message-ID: <8660a3a10701071705r4e9ce9a3ye189d4810c711071@mail.gmail.com> > > The list of supported options for the DS10L has a floppy/CD-ROM combo > drive which works nicely. > > Hans. > And costs more than what David Turner sells DS10Ls for. WWWebb -- Ajilon Consulting Site resident at Quest Diagnostics first.x.last@f$edit(contents of previous line,"COLLAPSE").com ------------------------------ Date: 8 Jan 2007 00:12:18 GMT From: Hans Bachner Subject: Re: SYSMAN suggestion: SET ENV /CLUSTER /EXCLUDE= Message-ID: JF Mezei wrote: > Tom Linden wrote: >>> SYSMAN> SET ENV/CLUSTER/EXCLUDE=(node1,node2) >> Unless you have a very large cluster, why is this easier than >> SYSMAN> SET E/N=(NODE3,NODE4) > Consider a shutdown procedure where one node wants to tell all other > nodes to dismount its system disk (you cannot do a DISMOUNT/CLUSTER of > your own node's system disk, and instead have to use sysman to tell > each node to dismount your system disk). Another case where this would be useful is one of my client's configuration - a cluster with four application nodes and a quorum node with very limited capabilities. It has its own system disk and does not even mount the cluster's shadow sets. When doing maintenance work or other configuration changes I often need to restart an application, (re)define logical names or other things on all application nodes but not on the quorum node. SYSMAN> SET ENV /CLUSTER /EXCLUDE=node would come very handy here. The suggestion to use logical names for node lists has problems when not all nodes in the cluster are running - the missing node is assumed to be a remote node outside the cluster which needs to be accessed over the network. Hans. ------------------------------ Date: 7 Jan 2007 23:31:20 GMT From: Hans Bachner Subject: Re: [ODS2_WIN32] Reading ODS2/ODS5 CDs/volumes on Win32/Win64? Message-ID: Hallo Peter, Peter 'EPLAN' LANGSTOEGER wrote: > What is the current status of reading/writing ODS2/ODS5 CDs/volumes > on Win32/Win64 systems (XP/Vista)? > > I remember having been able to read ODS2 CDs on Windows by downloading > the ODS2.ZIP from Hunters server (and Adaptec's ASPI package) many > years ago. > > But I'm currently unable to even get the least functionality (read > ODS2 CDs) on my XP Home SP2 notebook. > > ODS2_WIN32.EXE claims about ASPI problems. I thought, that ASPI is > already in XP (or maybe was installed by NERO, because Nero works with > my DVD burner) or is/was replaced by SPTI (SCSI Pass Through > Interface) eventually. I even tried to download and install ASPI (for > WinXP) from ADAPTECs side, but alas, still a no-go. works fine for me. I just copied WNASPI32.DLL from oen of the Nero folders into the directory where my ods2[_win32].exe resides. Here's the proof :-) E:\ODS-2 Reader>ods2 ODS2 V1.3 $> mount x: ASPI (Bus 1,ID 0) HL-DT-ST DVD-RW GWA-4080N 0C09 Removable 2048 %MOUNT-I-MOUNTED, Volume VMS073LP mounted on x: $> dir x:[kits] /size /date Directory x:[KITS] DECAMDS_073_KIT.DIR;1 1 6-APR-2001 16:01:42.68 DWMOTIF_VAX124_KIT.DIR;1 1 6-APR-2001 16:02:15.74 Total of 2 files, 2 blocks. $> This is Windows XP SP2 like yours. Schöne Grüße, Hans. ------------------------------ End of INFO-VAX 2007.015 ************************