What is the overhead in defining logical files and using them with 2 indexes. We have a file defined with 1100 bytes in the DDS. We would like to define individual fields by breaking them into 125 fields and provide Query/400 access. The file has 1 million records and gets updated only in the batch at night.
This is one that can't be answered directly, but I can give you some idea of what it will take. First, you will have an impact in disk space utilization from the logical files. I am not sure of the calculation in order to know the exact impact, but mirroring, RAID drives, length of key, uniqueness of the key and some other factors are involved in the size computations. As far as the speed, you have two areas to consider: During the day using query access, the queries may take a long time to run depending on the search, select and keys used to access the data. If the users will always be using the key (or partial key) of one of the logicals, the creation of the Open Data Path created by Query should take a shorter amount of time. If the queries must create a new Open Data Path (Sorting sequence and access to the physical data), then the existence of the logicals may be unnecessary. The impact to your nightly routine should be a small one, but if one of the logicals is keyed unique, or the keys are long, then the impact could be a big one. You will have to test with a small subset of the data in another library to determine what kinds of impacts there will really be.