[Info-vax] bound volume set limits

[abrsvc]

On Monday, July 3, 2017 at 10:38:01 PM UTC-4, David Froble wrote:
> mcleanjoh at gmail.com wrote:
> > On Tuesday, July 4, 2017 at 10:04:35 AM UTC+10, Hans Bachner wrote:
> >> Jan-Erik Soderholm schrieb am 03.07.2017 um 23:48:
> >>> Den 2017-07-03 kl. 19:19, skrev Baldrick:
> >>>> [snip]
> >>>>
> >>>> Scenario is this, VAX VMS 7.1 with just under a terabyte and a half of
> >>>> tape archived data of lots of relatively small files.
> >>>> [snip]
> >>>>
> >>> Maybe I'm just missing something here...
> >>>
> >>>  > 30 presentations of 50 gig.
> >>>
> >>> Would not 15 x 100 GB, or 5 x 300 GB work?
> >>>
> >>> But maybe 50 GB is at some limit for the OS versions used (?).
> >> I don't know the details, but "lots of relatively small files" might be 
> >> the key. The maximum number of files allowed on a disk depends on the 
> >> cluster size used:
> >>
> >> $ HELP INITIALIZE /MAXIMUM_FILES
> >> [...]
> >>       Restricts the maximum number of files that the volume can
> >>       contain. [...]
> >>
> >>       The maximum size you can specify for any volume is as follows:
> >>
> >>       (volume size in blocks)/(cluster factor + 1)
> >> [...]
> >>
> >> As increasing the volume size increases the cluster size, a bigger 
> >> volume doesn't buy you much in terms of allowed number of files.
> >>
> >> As the limit is applied to the individual volume set member (not to the 
> >> volume set), more files can be placed on a volume set than on a single 
> >> larger volume.
> >>
> >> Hans.
> > 
> > Good point Hans, but what's the file sizes and disk block wastage (i.e. unused blocks at the end of disk clusters), and what's the distribution of the wastage?
> > For example, if 90% of files use just 1 block and the disk cluster size is 7 blocks ...
> > 
> > Another option that might be possible is to swap many small files for fewer large files, but that will mean changes to the software, probably not only that which writes to disk but anything that reads those small files fom disk.
> > 
> > Running a museum piece like a Vax on VMS 7.1 isn't advised.  If support and maintenance costs are involved, a new integrity box will pay for itself inside 12 months (and maybe depreciable for tax purposes).  Or if it's a question of waiting for VMS on mass chipsets then maybe a simulator would be bridging solution.
> 
> Well, we don't know why he's running VAX/VMS, nor do we know if he's already 
> using an emulator.  Regardless, it's VAX/VMS V7.3 or earlier that determines 
> what types of devices can be handled.
> 
> Without that information, other configurations cannot be part of this discussion.
> 
> It seems to be conventional wisdom (CW) that for say a terabyte disk, or larger, 
> wasted space should not be a concern.  Maybe I'm just old school, but I don't 
> like that CW.
> 
> Nor do I have a clue what I'd do with a terabyte of storage, except for backup 
> save sets, which would be rather large files.  But that is just me.
> 
> Nor do I have much use for bound volumes.  I'd be looking hard for another 
> option.  Perhaps logical names representing multiple disks.  Perhaps other 
> solutions.

While it is generally true that bound volumes are no longer needed with larger disks, there are limitations to VAX that may require them as in this instance.

Lets focus on solving the posted problem rather than going down the rathole of larger disk availability.

Dan