sql load per node instead of all nodes at once

[kirankota79]

i have written an oracle table into dataset through 4 nodes. so the data is stored in 4 partitions. Is it possible to load the data through one node after the other instead of all 4 nodes at once? Bcoz i am thinking the error ora-00054 (resource busy) is causing because of this. No one is accessing this table nor it is opened in a toad session. Please suugest me about this...it is very urgent for me. It is happening to almost all the tables that i am loading and no one is accessing these tables.

[kirankota79]

on the output oracle stage...i am truncating and loading but i want to set the partitioning tab to "ordered". but i don't see this option...it is explained in the parallel job dev guide.

[ray.wurlod]

Because the Data Set is stored on four nodes it can only be read from those four nodes, typically with the same configuration file as the one with which the data were written.

You can bring the data together by forcing your Oracle Enterprise stage to execute in sequential mode, and using Sort/Merge as the collection algorithm to preserve any sorted order that exists in the Data Set.

[munch9]

If it is the multiple nodes executing at the same time causing locking issues you could try adding PARALLEL=TRUE to APT_ORACLE_LOAD_OPTIONS to allow multiple loads to run concurrently.

[girija]

Hi,

Its little bit confusing. What are you doing? Are you writing data into Oracle table from Dataset? If you writing data into oracle table and before that truncating the target table you need to synchronize your process, means truncate first and until and unless truncation finished no data will be inserted into the target table. Not only that, if you using bitmap index
in your target table, make them disable first and load the data.