>FLAME ON
Sorry, but I don't need no flame to express what I want.
>
>I wouldn't put a LISTSERV at DDAGSI1. I would start with putting a RSCSV1
>at DDAGSI1 -  something that doesn't reject files with  a sign-off record
>every 30 minutes.  DBNGMD21 can remain connected to RSCSV2,  and CEARN to
>RSCSV1 via  the more efficient  DMTVMB, and  everybody would be  happy. I
RSCS V1.3 is still available as node DDAGSI1O here in Darmstadt. If the
performance problem persists after wednesday ( a SE will check our GCS/RSCS
configuration if there can be tuned anything ) we can immediately rearrange
the routing.
>would also  put a  line monitor  there, so that  the line  gets restarted
>after a severe link  error. DDAGSI1 was down over 50% of  the time in the
>last week (at least, it appeared down in over 50% of the queries I sent).
On the DDAGSI1 side the automatic restart by RSCS is working, which restarts
the link immediately after going down because of an error or timeout limit
exceeded. In addition the NETCNTL monitor looks for all links every two
minutes, restarts in case of INACTIVE, notifies the OPERATOR in case of
ACTIVE, checks the links and queues for large files over a specified limit...
>Once we have a stable link to Germany, we could consider putting a stable
>LISTSERV there.
You should know that here is no more manpower at all for operating and
controlling a central server. If this can be done remotely by Jose, we (i.e.
the responsible people here at GSI Darmstadt) could think about putting
a server here.
>
>FLAME OFF
See above...
>
>  Eric
 
Happy EARNing...Udo Meyer
                GSI Darmstadt, Germany
                Manager Operations and Teleprocessing