[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference iosg::all-in-1

Title:	ALL-IN-1 (tm) Support Conference
Notice:	Please spell ALL-IN-1 correctly - all CAPITALS!
Moderator:	IOSG::PYECE

Created:	Fri Jul 01 1994
Last Modified:	Fri Jun 06 1997
Last Successful Update:	Fri Jun 06 1997
Number of topics:	2716
Total number of notes:	12169

2558.0. "%RMS-E-RNF, record not found on heterogenous cluster" by ROM01::FRISINA () Thu Mar 06 1997 17:47

    
    In a heterogenous cluster composed of two VAX systems, each of one with
    a system disk and SYSUAF file of its own, we get the error "%RMS-E-RNF, 
    record not found" when jobs are formatted.
    
    I think the problem is due to the fact the QUEUE_MANAGER checks the 
    username's record on the system that is managing the process at the
    moment.
    
    Unfortunately the systems must keep a SYSUAF file of their own, without
    adding any record referring to the users of the other node.
    
    Any hints about how to workaround this problem?
    
    Thank you
    
    Iolanda

T.R	Title	User	Personal Name	Date	Lines
2558.1		IOSG::PYE	Graham - ALL-IN-1 Sorcerer's Apprentice	`Thu Mar 06 1997 18:40`	6
	I'm sure you know that ALL-IN-1 is not supported on systems that do not have a common SYSUAF. Other things will probably break, e.g. the File Cabinet Server, in this scenario. Can you perhaps define queues that only run on the same node as the user that submits the job?
2558.2	run multiple queue MANAGERs	UTRTSC::SCHOLLAERT	Think before you shrink	`Mon Mar 10 1997 11:33`	14
	>Can you perhaps define queues that only run on the same node as the >user that submits the job? Won't work. Problem is the queue manager runs on one node by default. All entries for all queue's are validated by this queue manager. So the non-supported solution is running a queue manager on each node in the cluster. Check out with a local VMS guru. Cheers, Jan
2558.3	Unsupported by OpenVMS and ALL-IN-1 !!	IOSG::PYE	Graham - ALL-IN-1 Sorcerer's Apprentice	`Mon Mar 10 1997 11:52`	6
	<<<< So the non-supported solution is running a queue manager on each <<<< node in the cluster. Check out with a local VMS guru. You're right this is non-supported. There's recently been discussions about this in the VMS conference, where it was stated again that this wasn't allowed.
2558.4	Two possibilities	SHRMSG::HOWARD	Whoever it takes	`Wed Mar 12 1997 00:55`	7
	I suppose you could disable background printing. I'm not sure that you can do it system-wide. I don't even remember how to do it. What would happen if you defined a logical name to point one of the nodes to a different queue that only ran on that node, while OA$FORMATTER ran on the other node? That works for most things . . . Ben