Hi Dear Gurus:
We just upgrade Ingres 6.4 to OpenIngres 1.2 and we have weird
problems with the OpenIngres Communication process.
We have 3 OpenIngres servers (data) on Digital Unix V3.2G and 2
OpenIngres Clients (applications) on OpenVMS 6.2.
Our problem is that the OpenIngres GCC process ( the client
communication process) stop running mysteriously... in some cases
under high workload but in others cases under low workload. We really
can't understand the source of the problem.
We have spent to many hours trying to simulate the conditions in order
to reproduce the problem. For example, we simulated workload with 100
(or 200 or 300) batch process on one of the client machines(OpenVMS).
Each batch process executes a sql query on one of the servers
machines(DU). When we use a simple query the server machine runs
without problems. In this conditions, the machine doesn't have process
contention. When the query is more heavy (complex) the server machine
reaches 100% of CPU utilization (has a lot of sessions working
simultaneously) and under this conditions SOMETIMES the GCC client
process goes down, desappearing from the Operating System leaving no
messages at the ERRLOG.LOG file.
Should someone running OpenIngres with these or others platforms
willing to help us, just let me know, and I wil send more information
about the errors and OpenIngres configuration.
Thanks in advance for your time and help.
______________________________________________________________________
| |
| Carlos Dagnino Mardones System Engineer |
| e-mail: Carlos.Dagnino_at_consorcio.cl |
| Consorcio Nacional de Seguros-Vida |
| Santiago de Chile |
| |
|_______________[best regards from the end of the world]_______________|
Received on Wed Sep 03 1997 - 18:35:08 NZST