OpenVMS VAMP Board

The board was down for a day yesterday which was not planned

I had fitted a new SCSI card in my PWS shich had the effect of shoving my disks from DKA to DKB. Surprisingly few things fell over which I found remarkable (the power of logical names) however a few hard coded paths did not knock out Apache, SSL & MySQL.

I'm adding a new hard drive this week but will sign-post this one before I actually start unscrewing things.

Like I always tell my users, use the Disk$Label logicals, not $1$DKA100, in their code -- VMS always creates them. Then it doesn't matter *what* changes are made to the underlying hardware if the volume label stays the same. I moved my users from 20 RF drives to 24 1 & 2 gig RZ drives to 4 36 gig drives and no one skipped a beat (I created logicals that used all the old disk labels to point to the new drives when I went down to 4 big disks.) I then added a new system after that which shuffled some of the AlloClass numbers and the users never noticed the difference.

You are right of course.

I only needed to change about a dozen instances in sylogicals, systartup & sylogin and it was only layered products that were affected. I've now taken your advice and moved to using the label logicals.

Lovely design - move the entire system onto a different logical device and VMS doesn't skip a beat.

issinoho wrote:Lovely design - move the entire system onto a different logical device and VMS doesn't skip a beat.

Gotta love that "architected design". Imagine that, in an Operating System! Who'da thunk it? Obviously not bill g...

Alphaman wrote:[Gotta love that "architected design". Imagine that, in an Operating System! Who'da thunk it? Obviously not bill g...

Nor AT&T (MULTICS/UNICX ( -> UNIX)), Nor Linus T (Linux) and the rest of that gang (GNU).
Did you know, BTW, that though a lot of the architecture of BillG's system is based on VMS, the majority of the resulting software is directly derived from Xenix - his own Unix implementation? It explains a lot.....