I threw dispatch_io at a problem and was amazed at the result: superb performance! My only question is about memory usage.
In my case I am only doing streamed writes to a file. Since dispatch_io_write is async, there's the potential for me to supply data to dispatch_io faster than my hard drive can write it out which means that memory starts to stack up. What's preventing that from running away? I can generated gigabyte of data in mere seconds but it'll take the hard drive a lot longer to write it out, and I somehow want to know I need to stall data generation while dispatch_io catches up with actually writing to the file. The low and high water marks don't control how much data is in memory at any given time, but simply how when the io handler block is called based on the amount of data in memory. During writes, the dispatch_data_t passed to the io handler represents how much data remains to be written. One of my thoughts is to set the handler interval and check for the data size to be over my memory limit and if it is, set a flag and stall generation until it drops back below the memory limit which I would know the next time the handler is called and the data size is below the limit. I'm not sure if it this is a *good* strategy, but I can't think of anything else. -- Seth Willits _______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
