#4 How to Define and Load large databases with lots of tables

[Mike]

I have worked to create a large environment and have encountered numerous problems. I used v1.2 as it showed to be released. It drops connections all of the time and complains about missing tuples though it seems to have loaded the correct number of records in the tables. I loaded the tables from csv files an got lots of the following messages:
LOG: failed to find proc 0x7f6a7f638f20 in ProcArray
WARNING: pgstat wait timeout

There are 491 tables defined in this environment and 347 have data loaded for the application under test. There is about 30GB of data all together. On a number of occasions during the define of the tables and the load I received the following messages:
WARNING: unexpected EOF on datanode connection
WARNING: unexpected EOF on datanode connection
ERROR: sorry, too many clients already
ERROR: Failed to get pooled connections

Can anyone provide insight into this?