> I want to store some html files in a database having the following
> In order to retrieve the text, the title column would be
> searchable and indexed. I will not be searching on the text
> The average text length is ~4,500 bytes, but it is possible that
> the length could exceed 30,000 bytes.
> I am using perl (DBI) to connect to mySQL.
> I have a few questions:
> 1. What changes if any do I need to make to mysql in order to
> retrieve and store the text. My max_allowed_packet size is 1048576. Are
> there other buffers that are relevant if I am not searching on the text?
> 2. What is the buffer size in perl/DBI (as the client) for these purposes?
> Can it easily handle up to 60,000 bytes (both retreiving and inserting)
> without special handling?
> 3. Is the max amount of bytes that can be transferred via cgi using
> either the POST or GET method (ie., data in a TEXTAREA tag)
> 65,536 bytes?
> 4. Would a full-text search engine be better for my purposes or would
> mysql be better considering that I am not searching on the text.
1. MAX_ALLOWED_PACKET is the only you have to look for on the server side.
In your example it is set more than high enough (1M <-> 30K)
2. Most clients have a MAX_ALLOWED_BUFFER (or equivalent) size of 64K or greater.
This is the limit for retrieving data from the server. In your example high enough.
3. GET is limited (sorry, don't know the exact limit) when sending data from the client to
the server. POST was introduced, to overcome this limit. Therefore I would definitely use
the POST method.
4. It depends :)
For getting infos about using a relational DB versus filesystem look in the mailarchives
after BLOB diskussion of this thema.