On Wed, Aug 27, 2003, Kinkie wrote: > In fact, it's a common trend in current RDBMs: they're starting to favour > storage areas on fileystems rather than on raw devices. Linux has better > support for filesystem-based raw operations (O_DIRECT) than to raw devices > proper. > > BTW: O_DIRECT would be a good approach to our FS I/O work but: > - we need to be able to keep an object hot after reading it - the OS won't > do it for us anymore > - it only supports reading and writing on page boundaries, typically 4kb > (we'd need to be able to pad)
I don't really think we should even look at O_DIRECT until we've sorted out our own memory cache. The current "cache" isn't very flexible. I'd like to commit this change to squid-3 so I can start on some dirty/alternate rebuild logic. Duane, would you be willing to forward port your squid-2.5 COSS work to squid-3? I'm currently evaluating Squid-3 for a cache here and I'd like to test both epoll and COSS out. Adrian
