My two cents: I'm not convinced I ever felt any improvement on the desktop, even with horrendously fragmented system/boot volumes. A qualitative assessment for sure, but isn't the subjective experience the entire point?
I have seen notable improvements on file and database servers. With SSDs (including enterprise SSD caching solutions), defragmentation certainly is becoming ever more irrelevant. My role is primarily server administration though, so I have to defer to those with more extensive desktop experience. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Joshua MacCraw Sent: Thursday, August 8, 2013 3:25 PM To: [email protected] Subject: Re: [H] Defrag? On Aug 8, 2013 4:44 AM, "Thane Sherrington" <[email protected]> wrote: > > At 06:35 AM 08/08/2013, Joshua MacCraw wrote: >> >> No need to counter with stats, you partially debunked yourself! A >> linear copy test has no bearing on fragmentation's effects on real >> world random access. It wasn't 6 or 7 years ago, it was less than 4 > > > Ok, so you say my test is invalid. Maybe so, but if it is, why didn't you bother to mention this back when initial discussion took place? > Dunno, let me consult my diary from then, LOL. > And you still can't offer a repeatable benchmark. Usenet? How many people use that? Accpac for DOS also benefits from defragging, but not many people run that either. > How many indeed? Ask Quade how many copies of NBP are downloaded each month or Giganews if they have enough customers to justify 5 years retention. Point is it causes fragmentation & suffers noticeably poor performance as result. You said defrag is not beneficial in general and that's not true and not measured using linear benchmarks. FYI you know about MyDefrag because of me the last time this came up. > Steam? Ok, there are some gamers out there who haven't moved to SSD > who might find an improvement, but once again, not a really common application for must PC users. > Right, the average gamer is wasting big bucks on SSD's when 3TB EMD is so cheap! > Hibernation? How often are you doing that? Is the savings in time > from waking up and going to sleep twice a day worth the hours of defrag time? Yes it is & several times a day when I stop using it for more than an hour. No, I an not suffering using a netbook because it's not being used to do CAD or game on so it performs just fine. On it and my workstation I still have mutli-GB swaps which also benefit from being on a defraged drive. Hours? Sure, once every month or so as low priority or idle so no suffering there either. > > And frankly, if you're using a netbook, you don't care about > performance anyway. > Wrong & snobbish. > Show me a significant, repeatable, measurable improvement in office applications/web surfing usage, and I'll be happy to run the tests to see what sort of improvements are possible. Heck, I'd even be interested in improvement in boot times (not hibernation). You're tweaking the problem to be right. Fact is there are advantages to proper defrag done in zones with gaps and then maintenance defrags of those zones with several more common than not situations. If you add & remove large volumes of data on a regular basis the need is greater. So gaming, Usenet or torrents, music/video/pictures, and yes the lowly swap/hibernation that many portables use & some desktops. Personally I add & remove hundreds of GB a month, though I long ago sequestered most of that to a separate partition & zoned it properly. > > Here's an interesting thing: It's virtually impossible to get two defraggers to agree on what is fragmented and what isn't. Even MyDefrag, if you run it repeatedly, will move files each time to you it, suggesting that there is no completely defragged drive. Because mydefag does not do exhaustive defrag, because multiple passes produce varying landscape even so the files themselves DO end up defragmented (and in their zone) even if not totally sorted within the zone optimally assuming the correct logic and options are applied by the MYD script writer and sufficient free space to move in. > > T > >
