[Info-vax] Long uptime cut short by Hurricane Sandy

[Paul Sture]

In article <kdunhn$gek$1 at dont-email.me>,
 Stephen Hoffman <seaohveh at hoffmanlabs.invalid> wrote:

> On 2013-01-25 19:31:20 +0000, AEF said:
> 
> > I don't know. People used to brag of long uptimes here.
> 
> Yep.  I thought it was cool, too.  Then I thought of the implications 
> of what it meant.

I thought it was cool too, especially when my counterparts looking after 
Windows systems were rebooting so regularly.  The Windows NT 4.0 Server 
Resource Kit (aka Server RK) actually recommended in print that you 
reboot at regular intervals, with IIRC once a fortnight being considered 
a useful rule of thumb.

I changed my mind about long uptimes when I realised that in our 
environment where a lot of development was going on, regular restarts 
were a Good Thing to test startup and closedown procedures.  If 
something broke:

a) the changes were fresh in the implementer's mind
b) there were less audit logs to plough through if we needed to know who 
had made a change


The backup cycle is a consideration here too.  With a 
daily/weekly/monthly/quarterly/yearly backup cycle, the sooner you can 
identify something that needs restoring you have a better chance of 
nailing things down to the exact file versions where a change has been 
introduced

-- 
Paul Sture