Record-Not-Found when using SUBMIT/USER

[Rich Hearn]

Volker,

1st of all, thank you for the information. I see no way of entering points for this answer, but if I could, I would assign "10" - you're "spot on" for what I was hoping for.

I neglected to state that this involves a 2 node cluster (my apologies), but that did not stop you from providing an answer that included that scenario. Thanks again.

John,

In my particular case, I don't need all processes to switch, just any new ones. You provide a very interesting way to solve a problem that I, fortunately, don't have, but that's my fault for not being more clear originally - I apologize for that. I thank you though for it.

Background: Node1 was booted with an old UAF on it's sysdisk so we could get the common disk accessed & tested. Once that was done, Node2 was booted with the logical pointing to the current UAF (common disk). I'm finding I can't submit jobs to the queues I used to. Using the info from Labadie & Hein (earlier), I see my Node 1 is pointed to the UAF on the sysdisk (old). I want to change that to use the UAF pointed to by the logical (current) - Volker's answer does it quite nicely. I'm fortunate, in that, the only queue jobs executing are mine (others are holding), so I can afford to stop them should that occur - it's a bonus if they don't.

Hein,

I probably won't be writing anything to change the records for now however, I thank you for your answer from Sept 2003 (earlier response). Your current response merits some thought on my part - to understand how/why/when I might want to use it - worth the time. Thank you for giving it.

Volker,

Appreciate the clarity regarding the cluster failover vs the stop/id Thanks again!

Robert,

fyi: it may not be clear that I've "resurrected" an old note - I don't know if Rob will/will not respond. You raise an interesting point tho' - thanks for the perspective.


Thank you all for your responses and thoughts - as with VMS - they're elegant...
Rich

[Robert Atkinson]

As per thread