What determines if a bytea has 1 or 4 extra bytes?

All we need is an easy explanation of the problem, so here it is.

When inserting a BYTEA string in PostgreSQL the documentation says that such a string will occupy 1 or 4 bytes plus byte string length.

The question that I can’t seem to find the answer to is what determines whether 1 or 4 bytes are added and is this effect consistent or will it vary depending on some factor or other?

How to solve :

I know you bored from this bug, So we are here to help you! Take a deep breath and look at the explanation of your problem. We have many solutions to this problem, But we recommend you to use the first method because it is tested & true method that will 100% work for you.

Method 1

I believe the answer is in this header comment of heaptuple.c:

* Before Postgres 8.3 varlenas always had a 4-byte length header, and
* therefore always needed 4-byte alignment (at least).  This wasted space
* for short varlenas, for example CHAR(1) took 5 bytes and could need up to
* 3 additional padding bytes for alignment.
* Now, a short varlena (up to 126 data bytes) is reduced to a 1-byte header
* and we don't align it.  To hide this from datatype-specific functions that
* don't want to deal with it, such a datum is considered "toasted" and will
* be expanded back to the normal 4-byte-header format by pg_detoast_datum.

Note: Use and implement method 1 because this method fully tested our system.
Thank you 🙂

All methods was sourced from stackoverflow.com or stackexchange.com, is licensed under cc by-sa 2.5, cc by-sa 3.0 and cc by-sa 4.0

Leave a Reply