[Info-vax] Programming languages on VMS

[Bill Gunshannon]

On 02/01/2018 07:47 PM, Arne Vajhøj wrote:
> On 2/1/2018 7:24 PM, Bill Gunshannon wrote:
>> On 02/01/2018 07:08 PM, Arne Vajhøj wrote:
>>> On 2/1/2018 6:22 PM, Bill Gunshannon wrote:
>>>> On 02/01/2018 06:00 PM, Jan-Erik Soderholm wrote:
>>>>> Den 2018-02-01 kl. 23:37, skrev Bill Gunshannon:
>>>>>> On 02/01/2018 05:27 PM, Jan-Erik Soderholm wrote:
>>>>>>> Den 2018-02-01 kl. 20:42, skrev Bill Gunshannon:
>>>>>>>> Really.  So when the system disk dies who replaces the software
>>>>>>>> on it?  The PC weenie down the hall?
>>>>>>>
>>>>>>> You replace the failing disk in the mirror-set and HBVS rebuilds
>>>>>>> the failing disk. No software is "replaced".
>>>>>>
>>>>>> Who is "you"?  The secretary?  The Credit Union Manager?
>>>>>> The PC weenie down the hall who is lucky if he can install
>>>>>> a printer?
>>>>>
>>>>> Well, if you do have a system, I guess you do also have someone
>>>>> to fix small issues or someone to call in. Note that a failed
>>>>> disk in a mirror-set is not that urgent, in particular if it
>>>>> is an 3-disk mirror set. It is not a show-stopper and you
>>>>> can get it fixed at office times.
>>>>
>>>> But that was my point exactly.  It would take someone familiar with
>>>> not only VMS but the hardware that VMS runs on, which is not PC
>>>> class hardware.
>>>
>>> You don't think a normal bright guy can pull out a hot swapable
>>> disk and put another one in?
>>
>> Possibly.  But what about the commands needed to integrate the new
>> disk into the shadow set?  I have admined numerous VMS systems as
>> well as more than a dozen other OSes and I couldn't do it without
>> the documentation.
> 
> I most admit that I have zero experience with software/host
> based mirror.
> 
> Maybe that is a problem.
> 
> If it was hardware based mirror, then it would just work.

Even hardware  mirroring is not always as simple as you seem to
think.  I have had numerous systems where in order to "fix" a
RAID Array the whole system had to be taken out of service for
some rather ridiculous lengths of time. (This the reason my
strategy at the University in the latter years was mirroring
systems so that the entire data array could be replaced while
a broken system was repaired.)

bill