Diskeeper and Skyrim, questionable mix.

[CovertSlinky]

So, I use Diskeeper 12. Have for like...ever. Never really thought much about it's background service and what the intelliwrite function could be doing when gaming, never really had a reason to. But earlier this morning, after giving up practically, on why for the love of talos is skyrim ctd'ing at phantom and random moments without any noticeable cause, I have studied the papyrus log over and over for a couple of days to see really nothing that could justify it and I have concluded it to not be skyrim's fault. I have two PC's in the house, mine and the wifey's. Her computer is retarded, tis no joke, it's like 8 years old and still kicking, but for some stupid reason I could never get diskeeper to install right on her pc, and to be honest I gave up easily cause I hate that toaster. So, anywho, her pc no diskeeper, mine yes. So I stole her pc and setup skyrim on it to totally mirror mine, literally just copied and pasted everything I could over the network... yeah that was... slow. So, I then finally started playing on her pc, doing everything I normally do. I thought for sure her pc was just gonna straight puke all over me after a min or two of playing, mostly cause her's is so old and I really didn't think it could handle it. BIG F'in SURPRISE when it didn't, and then when I went an hour into the game, I started trying to FORCE it to crash, it was really pissing me off at this point. I was just like wtf???? how in the world could her pc be slapping mine in the face??!!!!?? SO....three hours went by, and I gave up, settled for just concluding that her pc was retarded and that is why it did what it did, and thats my story and I WAS sticking to it. Until 2 hours ago.....

 

Also to note, that her pc did have a substantially poorer frame rate (naturally) it averaged about 32 with a low of 19 and a high of 38...yuck!

 

I couldn't let it go.... I had to know, HAD TOO! PERIOD. So, I started thinking about ANYTHING I could that could possibly hinder the game, outside of the game. I looked and made sure my anti-virus and malware were actually shut off, no virus', no hard drive errors, no strange processes or memory leaks, etc etc. Till I got to looking over the services that were running and I gave diskeeper service a second thought, and wondered could it be? And I thought, well it does do all kinds of intelligent background procedures at times, and some are always, like intelliwrite. So I said fudge it, I'll just shut it off and see. My initial thought was that I would see some form of negative performance, because of the I-FAAST feature, that boosts file performance. Well, I was wrong, no negative, in fact, it was positive, skyrim loaded up 43 seconds faster (I timed it, several times). A noticeable amount of the stutter I would get while trying to move or look around right after loading (clicking on a door and thus loading a new cell), was gone, not all, but enough. And... no phantom ctd, idk if it's just coincidence or it really helped, I plan on doing more testing with it before I call it. But just thought I would share this and ask, if anyone else uses diskeeper and whether they know enough about it to know if it is linked or not? Any thoughts are welcome. I plan on posting back after enough deliberation on my part with my conclusion.

 

Regards,

Slink

[DoubleYou]

Looks like a fancy disk defragmenter in real-time. Can't imagine something like that could even help a modern pc.

[Aiyen]

Well it could be that the game simply does not get access to the resources it requires because it gets put on .... hold by that program. After all if just a single .nif file fails to load then the game will CTD. Also like Doubleyou says... in this day and age then it is really questionable if such utility software is even worth it. 

The lack of stutterings could also support this, since stuttering and micro stuttering is because the game is waiting for resources to load. In essence you are asking for too much data to be moved around relative to the available bandwidth in your computer. Using such software will reduce the overall bandwidth since the computer has to wait for it to perform its function before allowing other programs to get access. 

[CovertSlinky]

Very good point Aiyen. I had a feeling it had something to do with not allowing skyrim access when it requests it, and they can not be good for a game like this and then throw a couple hundred mods at it. This theory that is sounding more like a law, winds up changing things for me. The settings I have set enboost to would probably deserve a new look over. As far as Diskeeper being good, to be honest it really really really is that good of a defragging program. It not only takes away some of the manual time needed to keep up with  hdd defragmenting, by intelligently writing them to places on the disk that make sense and thus eliminates huge chunks of fragmenting before they happen. And it really is resource friendly. Idk, I have used it for a long time and have never run into a situation like this, but it all makes sense. So, I'll just have to make sure to shut it off when I play, no big deal. I also use another defragmenting program alongside diskeeper, piriforms defraggler. I have a paid license of it, but it is also a free utility, and very good imo for being free. Anyway, I digress. So to be safe just shut it off. Thanks for your inputs Doubleyou and Aiyen, I appreciate it.

[phazer11]

Yeah, I used to use Diskeeper. I even used Auslogics Professional, their real-time defrags should be disabled. Now (after doing a lot of mechanical drive research, in trying to recover data off my drives; I already knew a few ways but they were all descructive) I'm thinking about not using any of the professional defragmentation programs.

https://www.overclock.net/t/1405088/benchmark-should-i-defragment-my-system-which-defragmenter-is-best

[CovertSlinky]

Interesting read phazer. However, I STRONGLY disagree with that person's findings and recommendations about defragmenting programs. First, let me state the obvious, his test pc, even for 2013, is a freaking TOASTER. And he decided to test advanced programs that clearly state that you will see better performance on large capacity drives, using an 80 gig drive?!!?! That totally skews the results. Furthermore, he makes his statement about free space extents, which with professional defraggers like diskeeper and perfectdisk, are created more heavily, but more importantly small in average size for a reason. They are there to provide wiggle room for programs/files that have tendencies to bloat/increase in time. Like on my drive for instance, I have a total of 31,164 free space extents, but it comes in the form of an average extent size of 15.38 megabyte's! The largest extent is 246.83 gigabyte's, that's more than half of the free space left on my drive, 468.02 gigabyte's. The theory behind this is, that if the program creates small free space extents after a certain known file/s that continuously increase, it will eliminate some if not most of the need to keep defragging those file/s everytime they increase. Idk, I am no guru on the subject, but I know enough to know, that, that is terribly misleading. Just my two cents about the subject. 

[DoubleYou]

I do believe defragging can be necessary on regular harddrives. However, with the built-in one that runs on a schedule with windows, I've found very little need to manual defragging, unless you just got finished installing 50 programs like I did when I set up my new pc and fragmentation was getting high. Just checked my the Windows defragger, and it says 0 fragmentation. Checked Defraggler, and I got 2% fragmentation, and all the files it finds fragmented are merely temporary files from my browser and such, which really don't need defragmenting. This is on a 1 tb 5400 rpm drive running Windows 8.1.

[CovertSlinky]

Right you are, about the need. It really boils down to how far the user him/her self is willing to go and maintain the hdd. Tbh, the real time defragger I got, for me, isn't really a must have, it just winds up being more convenient. From my experience I have found the windows defragger to do it's job fine in regards to defragging a drive, nothing more nothing less. But it really needs to monitored often, cause if for whatever reason windows defrag stops (which I have had happen) or begins to decrease in it's ability to defrag the drive (which to I have had happen) completely, it will run away and get to the point where it can not fully defrag the drive. Windows from my experience typically has a hard time defragging large files with large amounts of fragmentation, and this more often than not becomes the root of windows defrag getting worse at it's job over time, especially large drives like yours and mine. But if you keep track of it like you are now, if and when this happens you can address it before it becomes a problem. And which that is what the "real" trade off from using a defragger other than windows is. But before I used diskeeper I just went solely on using windows and defraggler, windows was doing it job, and defraggler I only used to monitor information and statistics about my drive visually, and on occasion defrag a file/s that windows left behind. And now that I am sitting here typing this, and it's on mind, along with seeing points with the way of how I use my pc, getting in the way, just shelving diskeeper, and going back to that old chestnut. 

[phazer11]

Well of course it's a bit misleading. I actually meant to post the link to the defrag program. That thread is just the place I could find it lol (it also got me started doing my own tests, and I do really like MyDefrag btw)

 

Edit: Huh I guess the download link was removed. There were links to all of the defragmenters used in that test.

https://www.mydefrag.com/

[generalmx]

There are various things modern Windows (Vista and higher) does already in the background that can seriously conflict with a "real-time" defragger. Here's one benchmark for Windows 7, which shows some of the power of modern Prefetching and boot optimization: https://www.hofmannc.de/en/windows-7-defragmenter-test/benchmarks.html . I think I'll give MyDefrag a shot myself.

 

I'd also like to point out that, if you have a few bucks, refurbished SSDs can be purchased on eBay for ~$30 USD shipped, which can be used either alone or with Intel SRT to do SSD caching on compatible systems. You can also get the "Sandisk Readycache" SSD for ~$40 USD shipped on sale, which can work on most anything.

 

But the big question for Skyrim modders would be: how does this all actually influence Mod Organizer users? I'm sure the virtual file system adds some type of latency, but is it significant for mechanical vs solid state? Even if MO only creates the links in memory, it will still take non-zero time to parse the link, so that added time may give a slower drive more time to catch up.

[CovertSlinky]

Yeah, sorry if I made it sound like I was confronting you negatively about that link, was not my intention, also I didn't put it together that you were just trying to show me a download link, lol, my bad, totally flew over my he.... I mean, yeah download links were not there yup, huh, I knew that was significant, :p. I'll check that MyDefrag out though, cause it isn't the first time I have had it positively referred to me. And would be a good time to try it, cause I just un-installed diskeeper today, and went back to using windows and defraggler side by side. But if this winds up being as good as I am told then, I am down for "One defragmenter to rule them all", lol.

 

Ughh, man don't get me dreaming of ssd's man, lol, it's coming, it's coming darn it. I have been patiently waiting for that addition to my recently built pc, that I had to cut out of the initial build cause of needing to cut some corners budget wise. The cpu, psu, one more gpu were put on the back burner, at the beginning with intent to grab later. The ssd's, I thought I had enough for but didn't, so had to put them on the future plans list. So far I have acquired the psu I wanted but is not installed right now as it would be over kill, and now I am about two paychecks away from having the extra money for the ssd's, which was a toss up between the gpu for crossfire or the ssd's. So they are coming, finally. I'll have two plextor px-256m5pro extreme series ssd's, YAYYY, lol.

[phazer11]

Oh. To be honest. If you could (or you got an amazing deal on the plextor) I'd just grab me a Samsung EVO 250GB (hell if I didn't have to get these storage drives to replace my dying HDDs I'd buy me one too, just so I'd have more space, a dedicated drive for modded games like Skyrim and the Mass Effect's

M.E.U.I.T.M. texture pack). I'm running Corsair Neutron GTX 240GB's atm and they're really nice (I know people who have the EVOs and PROs though; enough to recommend them here and to Neovalen and William). Of course with SSD's you're probably going to get blown away even with a crappy SSD just because they're faster.

Here's some links you might find... interesting.

https://www.amazon.com/Samsung-Electronics-EVO-Series-2-5-Inch-MZ-7TE250BW/dp/B00E3W1726https://www.amazon.com/Samsung-Electronics-Series-2-5-Inch-MZ-7PD256BW/dp/B009NB8WRU/ref=zg_bs_1292116011_7https://www.tomshardware.com/reviews/ssd-recommendation-benchmark,3269-8.html

[generalmx]

I actually do not recommend the Samsung 840 EVO because it's one of the only SSDs which still uses 3-bit MLC (also known as "TLC") instead of the conventional 2-bit MLC --- 3-bit means less IOPS in certain circumstances (note that because 3-bit MLC is still technically MLC, you'll see the EVO called "MLC" in product pages when it's actually TLC). When you see the 840 EVO in any benchmark, that's with its "TurboWrite" caching feature enabled (as that's the only way it can keep up with traditional MLC), which significantly decreases its resilience to power-loss scenarios, as it's an added cache; often they'll add in its performance with RAPID Mode enabled to the benchmarks too, which is yet another cache. And all of that caching and it doesn't perform significantly better than the Crucial MX100, which is the "budget" SSD that ##hardware on Freenode usually recommends instead.

Edited August 10, 2014 by generalmx

[CovertSlinky]

Hey thanks for those tidbits, I'll give them a look at. I have always been a plextor fan, I have always been able to rely on them w/o regret. But, I'll give them a look at today, ain't nothing wrong with saving some bucks w/o having to sacrifice performance. But if what generalmx is true, than I don't think I'll be going that route. Anyone else able to relate to what generalmx is saying? I'll do my own research on it, but would be nice if anyone has working experience with them.

 

EDIT: Well after doing some brief looking, I think I am going to stick with the plextor's, I can pick them up for only $20 more than those samsung's, $149. The samsungs are a sata 3, plextor sata 6. The plextor offers a 5 year warranty, samsung offers 3. The specs that each company posts, are close, plextor has a higher random read/write, the samsung only beats out the plextor in sequential write speed, 460mb compared to 520mb. The sequential read is the same for both, 540mb. However, samsung doesn't make it known what there dram cache size is. The plextor has a 512mb ddr3 cache size, considering the specs that samsung is listing, I would have to think it's something close to the plextor. Also, that benchmark from tomshardware listed the m6s, which is slower than the m5pro, from what I am reading, the m6 series claims to be more efficient than the m5 series by what looks like software boosting or something, idk, cause the specs that plextor states is that the m5 series is not only cheaper but faster.

Edited August 10, 2014 by CovertSlinky

[generalmx]

I think you're confusing SATA-III with SATA 3Gb/s --- almost all modern SSDs are SATA-III, which is 6Gb/s; there is no "SATA6" interface. It goes:

 

SATA Revision 1.0 - 1.5Gb/s

SATA Revision 2.0 (aka SATA-II) - 3Gb/s

SATA Revision 3.0 (aka SATA-III) - 6Gb/s

 

Note: Each revision is fully backwards compatible, so a SATA-II port will work with a SATA-III drive and visa-versa, but not get the full speed of course.