[Info-vax] probably a simple issue...

[Stephen Hoffman]

On 2014-02-17 15:24:04 +0000, cyberdefensedave at gmail.com said:

> It turned out to be a far simpler problem, and I think I've actually 
> got this whole Decnet thing sorted out now.  Volumes mount, monitor 
> cluster functions... whew.
> 
> Apparently 1.2 had not managed to complete its reboot.  I wandered down 
> to the datacenter and realized that the captive accounts were being 
> refused login as well, which I have seen happen in the past when 
> something gets stuck during a shutdown/reboot.  Forced a dead sergeant, 
> booted and all is well.  The renumbering of the node to 1.2 seems to 
> have finally stuck!

One of the more common triggers for a stuck bootstrap in OpenVMS is the 
queue manager.  That software is too dumb to report an error and 
continue if a command to start a queue is invoked when the cluster 
member associated with that queue isn't available.  There's no clean 
way around this either — well, other than ensuring the hosts are 
available — as DCL only has the most hackish and limited support for 
command timeouts.



-- 
Pure Personal Opinion | HoffmanLabs LLC