We use SRS6 on Linux for DDBJ, which includes daily updated files.
I made indices for it. These became nearly 1.7G as a maximum size.
On linux(ext2) each one file was treated until 2G.
I knew SRS6 has a function, which can do parallel indexing.
I tried using "parallelType:fileSize partSizeKb:500000" in ddbj.i file.
It created indices with parallel, however, at last it put partial
Does anyone achieve the same as our aim, which we want to
control indices separetely less than 2G on Linux, even though
database source files will be larger and larger in near future.