There are some use cases that can make it use more memory than you
might think. It buffers the results in memory, so a large result set
can cause issues, this can be compounded if you are running a large
number of threads since each thread will have it's own buffer pool.
While dbslayer 'frees' it's memory, like almost all programs it
doesn't call sbrk to return memory to the OS while it is still running
- so once allocated the memory will stay with the process until the
process is terminated - but this is standard across all programs
running on unix.
How big are the result set you are return? Any other details? Platform
specific might be helpful.
d
> --
> You received this message because you are subscribed to the Google Groups "dbslayer" group.
> To post to this group, send email to dbsl...@googlegroups.com.
> To unsubscribe from this group, send email to dbslayer+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/dbslayer?hl=en.
>
>