Ansicht
Dokumentation
03731 - Problems with RFC destination for Vertex
SUBST_MERGE_LIST - merge external lists to one complete list with #if... logic for R3up TXBHW - Original Tax Base Amount in Local CurrencyThis documentation is copyright by SAP AG.
Problems with RFC destination for Vertex
Hi Volker,I had previously tried giving *allobj authority to PRD09 for the
Vertex SQL collection but the problem arises is that Vertex uses a
parameter file where you must code the used id to use during the call
to the SQL collection. However, the job that gets spawn on the
AS/400 from the RFC call runs under a different user id. You end up
getting an SQL7022 error that states the user specified in the
connect is not the same as the user in the current job.
Charlie Wright, who is on this forum, provided me with useful
instructions he had received from IBM to set this RFC destination up
as a "registered" destination. I performed this configuration
yesterday and this worked very well. I do not know why Vertex does
not send out instructions to set up this RFC destination this way. My
response time to test this connection under SM59 went from 1200ms
down to 40ms after making this configuration change. Thank you to
both you and Charlie for providing me with this recommendation. I
find this forum to be very useful and informative.
As for the second instance, in the past I had run a separate instance
on our S30 that handled the majority of the batch load for our PRD
system. On this S30, I also run our DEV and TST systems. Our SAP
development team has recently been complaining about the response
time so I decided to try to move this batch PRD workload back over to
the 740 where our primary dialog users run. We have about 13gig of
memory on the 740, so I thought I would create a separate instance
and put it in a separate memory pool so I could control the impact of
the background jobs that run. Our average dialog times for about 250
concurrent SAP users is about 370ms so I need to maintain their
response time. I have tried in the past to move this batch workload
under the same instance but I have run into issues with parallel
processing jobs taking over the instance. I have not had much luck
getting RFC quotas to work. We are currently on 45B but we are
looking to move to 4.6 or whatever is available the 2nd quarter on
next year. I am also hoping to soon consolidate our S30 and 740 into
a new 840 and use LPAR.
Kind regards,
Russ Cook
--- In SAP on System iZy..., "Gueldenpfennig, Volker"
<volker.gueldenpfennigZs...> wrote:
> Hi Russ,
>
> wouldn't it be possible, to authorize more than 1 user for access
to the
> SQL-collection ? When it would be a real OS/400 authority issue,
you culd
> perhaps use GRTOBJAUT or as test grant *ALLOBJ for PRD09.
>
> On the other hand there is the question, if you can't run vertex as
> RFC-server instead of RFC-client, that registers itself at the
gateway. Then
> there is always the vertex-user (or when you prefer PRD02 as well)
the one
> that started the RFC. Then you'd have to reconfigure SM59 from
starting to
> register.
>
> As last question, I would like to ask where for you need the second
> appl-server ? Is it perhaps possible, to increase the first
instance ? (when
> you are on 4.6x you can increase the Extended Memory as high as you
want,
> when the kernel patch is high enough.
>
> Regards
>
> Volker
Durban Tours - Südafrika Safari
RFUMSV00 - Advance Return for Tax on Sales/Purchases Addresses (Business Address Services)
This documentation is copyright by SAP AG.
Length: 3772 Date: 20240426 Time: 214204 sap01-206 ( 4 ms )
