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

[cyberdefensedave at gmail.com]

On Sunday, February 16, 2014 12:46:58 AM UTC-5, JF Mezei wrote:
> On 14-02-15 20:29, wrote:
> 
> 
> 
> > Things seem to have gone well.. I simply reconfigured one node as a satellite.
> 
> 
> 
> Did you make sure to move all node specific configuration files to
> 
> SYS$SPECIFIC and all common config files to SYS$COMMON ?
> 
> 
> 
> [ ala disk:[SYSx.SYSEXE]  and disk:[VMS$COMMON.SYSEXE]  ?
> 
> 
> 
> In a cluster, each node must still have its own distinct system root
> 
> with the node specific files in that root.
> 
> 
> 
> This is a bit far now, but for separate nodes:
> 
> 
> 
> on NODE-A:
> 
> 	NODE-A is defined as executor
> 
> 	NODE-B is defined as an end node.
> 
> 
> 
> on NODE-B
> 
> 	NODE-A is defined as end node
> 
> 	NODE-B is defined as executor
> 
> 
> 
> I can't remember if DECNET can be in the common directory or if separate
> 
> decnet configs are needed.
> 
> 
> 
> If you share the authorization and proxy files, then the decnet
> 
> authentication is shared and you need to "merge" the 2 distinct ones
> 
> into a common one.
> 
> 
> 
> for instance nodeA::user1 maps to nodeB::user1
> 
> and nodeB::user1 maps to nodeA::user1
> 
> 
> 
> You can use node wildcards for the source, but you need separate entries
> 
> for each node.
> 
> 
> 
> *::user1 maps to node-A::user1
> 
> *::user2 maps to node-B::user1

Hmm..  I'm not sure on the answers here (or how to check) but I suspect that the answer is "Yes" since I am now booting and doing all processing from a shared system disk on host 1.2.  However, I did manage to sort out the disk mounting issue (neglected to run disk sharing service on node 1.3) and that node is simply being used to back up data.