On Fri, 2012-06-22 at 18:47 -0400, Sam Varshavchik wrote: > When that happens, the winner's going to be whoever gets out of the way > faster, and doesn't diddle around too long, before sending your bytes to the > disk, and doesn't waste its time getting them back to you. Because, diddling > around, and trying to optimize your file layout, and making sure stuff isn't > fragmented on disk; all of that, is not going to be as important as it's > used to be, that's where I think we'll end up soon.
I can see this happening for systems that more less access their super fast IO via their CPU bus, in particular once people start ditching disk access protocols and access SSDs more like RAM is accessed (e.g. do something like what Fusion IO does). But on the other hand, more and more storage is being accessed via some network channel, and whereas throughput for a single IO request might not be such a constraining factor any more in the future (so that things like zipping your messages become questionable), what about the IO latency, i.e. the delay and resource usage related to the number of requests rather than their total size? I mean, sure the manufacturers of network equipment strive to reduce IO latency these days, but don't you think the number of IO requests per user request is going to remain an important factor in determining response time and scalability in IMAP style applications once you factor in networked storage? ------------------------------------------------------------------------------ Live Security Virtual Conference Exclusive live event will cover all the ways today's security and threat landscape has changed and how IT managers can respond. Discussions will include endpoint security, mobile security and the latest in malware threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/ _______________________________________________ Courier-imap mailing list [email protected] Unsubscribe: https://lists.sourceforge.net/lists/listinfo/courier-imap
