Re: Mounting disks during STARTUP

From: Carl Bennett (carltbennett_at_yahoo.com)
Date: 02/15/05 

Date: Tue, 15 Feb 2005 04:19:27 GMT

I tend to agree with David...

Unless I am doing a quorum disk, (which gets stuck in SYLOGICALS so that I
can set up the roots for SYSUAF, QMAN$MASTER, etc... ), I like to mount
disks under a separate com file near the end of SYSTARTUP. It just makes it
easier if you ever have to get in and cut the startup off before it
finishes...

We also use a source file that gets opened and read for instructions on how
to rebuild disks at startup... makes it really handy for all of the KZPAC's
that we've been retiring lately to only have one file to update when the
ultra-scsi comes in...

SYCONFIG would be ok, if it was just one or two systems that I got to be in
front of all the time, but I couldn't imagine trying to talk a night
operator through editing that file when the pager goes off at 2 am...

carlbennett<>users.com

"David J Dachtera" <djesys.nospam@comcast.net> wrote in message
news:42115708.BE888AD2@comcast.net...
> John Brandon wrote:
>>
>> helbig@astro.multiCLOTHESvax.de wrote:
>> > > The reason to use SYCONFIG is to keep the startup strategy in tact.
>> > Can you elabourate on this last sentence?
>>
>> As I tried to explain in the first thread - each startup routine has its
>> purpose. I guess I am expanding on the intent of SYCONFIG by making it a
>> disk
>> startup procedure - right or wrong - I just got it in my head that
>> SYCONFIG is
>> the right procedure to mount the disks.
>
> I try to keep as much of the "user-land" stuff in SYSTARTUP_VMS and
> later. I like to do page/swap files in SYPAGSWPFILES, but some of my
> application vendors have a problem with this (they want to retain
> control of the environment).
>
> Also, during SYCONFIG, as you noted, that actual configuration phases
> are somewhat incomplete.
>
> I interpret SYCONFIG as good place for GK devices, MG devices and other
> stuff that I'd want to have even when I choose to skip SYSTARTUP_VMS,
> but still do a full VMS startup.
>
>> helbig@astro.multiCLOTHESvax.de wrote:
>> > I have a procedure called MOUNT.COM which resides on a disk which is
>> > not
>> > a system disk for any node in the cluster (actually, it is the user
>> > disk). This also contains SYSUAF and other such files. The idea is
>> > that these should be common across the cluster, so there is no point in
>> > having them on a) just one system disk or b) maintained in multiple
>> > copies on several system disks. I mount all disks but 2 here (see
>> > below).
>>
>> I use a similiar concept - I called mine DISKS$MOUNT_STARTUP.COM.
>> However I do
>> not hard-code my devices. The procedure reads a text file
>> (DISKS_DATABASE.DAT)
>> that contains pertinent information for each device (device name, label,
>> logical, and shadowset.) The procedure will perorm checks on the device
>> prior
>> to mount and reports back any problems prior and after the mount.
>
> I tried to do something similar, but it never evolved into what I
> wanted. I posted here very long ago (about six years) about a facility I
> had dubbed "Shadow-Set Constituency Knowledge Base" (SSCKB, for short),
> but other posters immediately started asking questions that seemed
> inappropriate to something written in DCL to manipulate and use data
> about which volumes belong to which shadow-set, what the correct MSCP
> host (of FC path) should be, where appropriate, etc. I've kept to myself
> about it since. There's even a CHKDSK procedure to flag such things as
> missing shadow-sets and/or members, devices with the wrong MSCP host,
> shadow copies and/or merges in progress, etc.
>
>> I created a CLUSTER$DISK that contains all my common startup routines.
>> The SYLOGICALS and SYSTARTUP_VMS reference this disk for the startup
>> routines
>> and executes them.
>
> Do you also isolate such files as the UAF, RIGHTSLIST, NET*PROXY, the
> queue database, and so on?
>
>> So we have similiar concepts in startup routines. Just different names
>> and
>> variants on hard-coding (I hate it when I have to hard-code).
>
> Same here, but that's what I encounter the most out here in the field.
> Cut-and-paste, redundant code - difficult or impossible to maintain.
>
>> Ken Fairfield [my.full.name@intel.com] wrote:
>> > Is there a reason you don't want to wait until SYSTARTUP_VMS?
>>
>> I feel that one routine should be responsible for mounting all your
>> disks. I
>> view disks no differently than CPU or Memory. Therefore my disks should
>> be
>> mounted very early in the startup process.
>>
>> Mounting them in (say) SYCONFIG I avoid having to put the same mount
>> mechanisms
>> in SYLOGICALS, SYPAGSWPFILES, SYSECURITY, SYSTARTUP
>> Why not just have them in ONE procedure?
>>
>> It is easier to manage. And if I happen to shuffle my startup disks
>> around it
>> will be transparent to the startup - as long as I do not change the
>> volume
>> label then I am in good shape.
>
> How 'bout those startup-time scenarios where you don't want the disks to
> mount? For example: last weekend, I did a system migration from the old
> datacenter to the new where I didn't want the second shadow-set members
> to MOUNT until the system had been up for validation and we had decided
> to forsake any back-outs. The shadow-copies took about 36 hours to
> finish up totally, and I expected as much; so, I didn't want to start
> something I'd have to abort later on if things didn't go right and we
> had to back out of the move.
>
> --
> David J Dachtera
> dba DJE Systems
> http://www.djesys.com/
>
> Unofficial OpenVMS Hobbyist Support Page:
> http://www.djesys.com/vms/support/
>
> Unofficial Affordable OpenVMS Home Page:
> http://www.djesys.com/vms/soho/
>
> Unofficial OpenVMS-IA32 Home Page:
> http://www.djesys.com/vms/ia32/
>
> Coming soon:
> Unofficial OpenVMS Marketing Home Page

Relevant Pages

  • Re: Mounting disks during STARTUP
    ... but still do a full VMS startup. ... > to mount and reports back any problems prior and after the mount. ... about which volumes belong to which shadow-set, ... > I feel that one routine should be responsible for mounting all your disks. ...
    (comp.os.vms)
  • Re: xp Media Center wont boot up
    ... If you have no disks there's not much you can do. ... If this is an OEM you can request the recovery media. ... >> happens" when trying the other startup options, ... >> need to perform a repair install with your XP install media. ...
    (microsoft.public.windows.mediacenter)
  • Re: Prevent internal volume automount?
    ... the correct systax for the startup item would be much appreciated ... The Apple article doesn't mention OrderPreference but even some of the ... Requires - Disks ...
    (comp.sys.mac.system)
  • Re: chicken and egg: shadowing and licenses
    ... > On my systems, I mount all disks, be it shadow sets or plain disks, ... The cluster wide logical name LMF$LICENSE points to this file. ... > And I get no license warnings during startup! ... I mount the cluster-wide disk from SYLOGICALS.COM and mount other disks ...
    (comp.os.vms)
  • What process might be running & how do I stop it?
    ... missing so I downloaded the file from Microsoft to create a set of startup ... When I extract the files & start making the disks I ...
    (microsoft.public.windowsxp.help_and_support)