[Gmod-schema] Breaking the 1Gbase sequence length barrier

[Joe C. <jwc...@lb...>]

Well, today we broke the limit with a chromosome longer than 1 Gbase  and I tried to insert it into a chado record. Yes, the postgresql text field is unlimited in length, but there is a max of 1 Gbyte in a string buffer that is used for operations. So I cannot populate that field.

I’m curious if anyone else has run into this issue before and what they’ve done about it. It seems that the thing to do is to separate the residues field from the feature table and divide residues into hunks of a more manageable size.

The next milestone is when we break the 2G limit and need to make fmin and fmax in the featureloc table bigints. That, however, will be an easier fix.

Joe

(by the way, there is some info on large chado tables here https://www.mcgeeandco.com/products/chado-large-table-lamp <https://www.mcgeeandco.com/products/chado-large-table-lamp>)