ciaobaby Posted November 2, 2012 Report Share Posted November 2, 2012 What you are saying is utorrent version 2.2.1 change with great success the hardware configuration to match all the computers it is used on.Nope I'm using 3.2.2 and have adjusted my cache, queue and bandwidth settings to suit how my system is set up.First off, as in Rafi's "Best Practice tips" have any of you set your cache size manually?Start off at 128 MB and if you have a fast connection go up to 256 MB to reduce the disc writes. cause i didn't configure anything in 2.2.1 That's because you are running an OLD client that was developed for slower hardware so it's NOT likely to have problem with any current hardware. Link to comment Share on other sites More sharing options...
boomman Posted November 3, 2012 Report Share Posted November 3, 2012 I experience ~15 second bursts of my normal speed ~4MBps, other times it's down to 100 kbps or less.HDD gets writes for 32 MBps, there is no way I'm downloading that fast.I use windows cache with all utorrent caching disabled. The bug is in utorrent.http://i.imgur.com/P7bil.jpg Link to comment Share on other sites More sharing options...
rafi Posted November 3, 2012 Report Share Posted November 3, 2012 I experience ~15 second bursts of my normal speed ~4MBps, other times it's down to 100 kbps or less.HDD gets writes for 32 MBps, there is no way I'm downloading that fast.I use windows cache with all utorrent caching disabled. The bug is in utorrent.http://i.imgur.com/P7bil.jpgYou can either do what *you* want, or do what *we* recommend... http://goo.gl/yPZLJ Then, come complain about it... Link to comment Share on other sites More sharing options...
boomman Posted November 3, 2012 Report Share Posted November 3, 2012 You can either do what *you* want, or do what *we* recommend... http://goo.gl/yPZLJ The come complain about it...Riight. Out of millions Win32 applications utorrent is the only precious little princess that has unique unfixable issues (even on latest OS version) with windows disk cache that is also used by thousands of enterprise applications as well. These unfixable problems cause the need to reinvent the wheel. Must be masons or secret government did it, to slow down file transfers.Far more possible is that you mess up several winapi file handling call parameters and it behaves like a drunk horse as a result. The article here written by people that actually know windows architecture explains what is wrong with magic utorrent cache: http://habrahabr.ru/post/107637/Alright, I enabled your magic cache which seems to have alleviated problem somewhat. The fact that you can't handle standard windows API is somewhat sad though. Link to comment Share on other sites More sharing options...
ciaobaby Posted November 3, 2012 Report Share Posted November 3, 2012 If you run uTorrent on an operating system that is DESIGNED and OPTIMISED for fast disc tranfer rates ie a SERVER version it won't be a problem.HOWEVER the "desktop" versions of Windows are configured for application performance and disc activity is a infrequent operation.BitTorrent clients work the other way around and disc activity IS a continual process and application performance is a background task. So you have to give a little helping hand so that the much faster memory can be utilised more and disc activity reduced.It is NOT rocket science! Get over your lack of knowledge about how computers and applications really work and READ THE F***ING MANUAL firstI'll give you a clue!Press Help -> click the 'Contents' tab -> click on FAQ -> then Trouble shooting -> Finally READ "Why do my torrent jobs grind to a halt ...." Link to comment Share on other sites More sharing options...
rafi Posted November 3, 2012 Report Share Posted November 3, 2012 which seems to have alleviated problem somewhat. Glad to hear that. I'm not disputing the fact that the default settings could have been set in a way more suitable for high speed connections, and lower grade hardware... The devs seems to be aware of it too, and are revising the I/O system in v.3.3 , which is available to you as well for testing (and I do NOT consider it usable yet) that is also used by thousands of enterprise applications Even if so, how many of them are free? And you are welcome to try my settings file too... who knows, maybe it will do you some more good... Link to comment Share on other sites More sharing options...
boomman Posted November 3, 2012 Report Share Posted November 3, 2012 Even if so, how many of them are free? "it's free so you can't say shit" is not really an argument.I welcome I/O system overhaul in next version.I'm not sure who ciaobaby's rant was directed towards, but I am very much interested to hear about how desktop versions of windows are not "DESIGNED and OPTIMISED" (lol) for fast disk access. The common knowledge tells me the only difference between server and desktop versions of windows in regards to memory and disk performance is prioritization of background processes and lack of SuperFetch service on server. Link to comment Share on other sites More sharing options...
ciaobaby Posted November 3, 2012 Report Share Posted November 3, 2012 Common knowledge? HMMMM? That's a bit like common sense isn't it.Everyone thinks they have it, but nobody shows the slightest sign of ever using it!So you don't think that an operating system which is designed SPECIFICALLY for the tasks of reading data from hardrives or any other storage medium and delivering it in response to multiple user requests isn't configured to handle disc access more efficiently than a "desktop" OS.Although with Microsoft it is actually done the other way around. The Windows desktop OS has limitations built in to it with "hardwired" connection limits and "slowdowns" on disc access to prevent people actually using it as a fileserver, back in the days of NT4, these settings were in a registry key and you could 'unlock' NT4 Workstation and run it as a fileserver. M$ put a stop to that when Windows 2k made it's appearance. lack of SuperFetch service on server.And why pray tell, do you think it is that "SuperFetch" and it's "PreFetch" predecessor are not on server versions, but are on the desktop OS versions??? Link to comment Share on other sites More sharing options...
boomman Posted November 3, 2012 Report Share Posted November 3, 2012 So you don't think that an operating system which is designed SPECIFICALLY for the tasks of reading data from hardrives or any other storage medium and delivering it in response to multiple user requests isn't configured to handle disc access more efficiently than a "desktop" OS.It is the same OS. They have the same kernel. They have almost identical API. There is no difference between them. You appear to think there is some, while not providing any evidence. Connection limits are not related to disk access at all.And why pray tell, do you think it is that "SuperFetch" and it's "PreFetch" predecessor are not on server versions, but are on the desktop OS versions???Because desktop users launch the same utilities every day? And servers handle a lot of variable data? Link to comment Share on other sites More sharing options...
ciaobaby Posted November 3, 2012 Report Share Posted November 3, 2012 That's correct!Because a server OS is setup for quick and efficient DATA TRANSFER while a desktop OS is setup for running applications on. Link to comment Share on other sites More sharing options...
boomman Posted November 3, 2012 Report Share Posted November 3, 2012 So, no actual differences aside from superfetch. I'm glad you agree. Link to comment Share on other sites More sharing options...
Kitsoran Posted November 7, 2012 Report Share Posted November 7, 2012 How???You are the ones complaining about it' date=' I don't have any such problems because I [b']READ properly and UNDERSTAND how things workyour denial seems to be very strong... but i'll explain anyway.it'll be easy:If 2.2.1 works fine for everyone and versions > 2.2.1 does not. its a code problem.thats it.Things are rarely that simple, and the complexities of what really goes on and why it aint necessarily the clients fault often need to be brought up, but that doesn't mean µTorrent is entirely innocent: http://blog.libtorrent.org/2012/05/windows-disk-cache/ Link to comment Share on other sites More sharing options...
boomman Posted November 7, 2012 Report Share Posted November 7, 2012 your denial seems to be very strong... but i'll explain anyway.it'll be easy:If 2.2.1 works fine for everyone and versions > 2.2.1 does not. its a code problem.thats it.Things are rarely that simple [...]No' date=' things are really that simple. If bug started to appear from certain version, the code change is the reason.From the link you provided it seems that developers started reading manuals on functions and parameters they are using, that is a great progress; however the problem that's been described in the beginning of an article has already been solved in Vista and up: I/O Prioritization in Windows Vista. Link to comment Share on other sites More sharing options...
boomman Posted November 7, 2012 Report Share Posted November 7, 2012 your denial seems to be very strong... but i'll explain anyway.it'll be easy:If 2.2.1 works fine for everyone and versions > 2.2.1 does not. its a code problem.thats it.Things are rarely that simple [...]No' date=' things are really that simple. If bug started to appear from certain version, the code change is the reason.From the link you provided it seems that developers started reading manuals on functions and parameters they are using, that is a great progress; however the problem that's been described in the beginning of an article has already been solved in Vista and up: I/O Prioritization in Windows Vista, so that manual caching thing all Utorrent Defenders here recommending is useless. Link to comment Share on other sites More sharing options...
Kitsoran Posted November 7, 2012 Report Share Posted November 7, 2012 If bug started to appear from certain version, the code change is the reason.It's hardly simple when (some) bugs are really caused by external factors interacting poorly with new versions of µTorrent. Some software injects code into other software with the goal of altering behavior, sometimes poorly. Indexers and scanners can lock files and cause poor resource utilization, and firewalls simply block the new exe at times. Link to comment Share on other sites More sharing options...
Kitsoran Posted November 7, 2012 Report Share Posted November 7, 2012 If bug started to appear from certain version, the code change is the reason.It's hardly simple when (some) bugs are really caused by external factors interacting poorly with new versions of µTorrent. Some software injects code into other software with the goal of altering behavior, sometimes poorly. Indexers and scanners can lock files and cause poor resource utilization, and firewalls simply block the new exe at times. Link to comment Share on other sites More sharing options...
ciaobaby Posted November 7, 2012 Report Share Posted November 7, 2012 So, no actual differences aside from superfetch. I'm glad you agree.well that maybe your interpretation but there is more than just "superfetch" being the difference be a client OS and a server OS. But if you want to take an entirely simplistic and meaningless viewpoint that is your perogative. Link to comment Share on other sites More sharing options...
Kitsoran Posted November 7, 2012 Report Share Posted November 7, 2012 No, things are really that simple. If bug started to appear from certain version, the code change is the reason.Completely ignoring when bugs are really caused by external factors interacting poorly with new versions of µTorrent. Some software injects code into other software with the goal of altering behavior, sometimes poorly. Indexers and scanners can lock files and cause poor resource utilization, and firewalls simply block the new exe at times. Link to comment Share on other sites More sharing options...
rafi Posted November 19, 2012 Report Share Posted November 19, 2012 More results related to the issue are here:http://forum.utorrent.com/viewtopic.php?pid=694936#p694936 Link to comment Share on other sites More sharing options...
infusednz Posted January 24, 2013 Report Share Posted January 24, 2013 Some real retards here. Latest versions are fucked. I run this on Win2k8 R2 using iscsi bonded adapters using balance-rr. 2.2.1 is excellent. Excellent being I can push through 90mb/s while hashing torrents.Anything after that is a pig. The cache is fucked. 3.2.3 I can download at 8mb/s before I get disk io problems and disk trashing. Going back to 2.2.1, problem is fixed.Probably my slow hardware and disks eh?But yeah, keep blaming the users. Link to comment Share on other sites More sharing options...
DreadWingKnight Posted January 24, 2013 Report Share Posted January 24, 2013 Oh hey, sequential benchmark.Shame that torrenting is NOT sequential. Link to comment Share on other sites More sharing options...
Firon Posted January 24, 2013 Report Share Posted January 24, 2013 hdtach is useless. try this: http://crystalmark.info/software/CrystalDiskMark/index-e.html Link to comment Share on other sites More sharing options...
midbie Posted February 6, 2013 Report Share Posted February 6, 2013 Try this, at the Speed tab in the lower panel, select Disk Statistics, change the resolution to 30 sec (larger interval may work too). It sounds silly, but it works for me, i dont have the fxxking annoying disk overload problem for few months. Link to comment Share on other sites More sharing options...
BigEd1 Posted March 25, 2013 Report Share Posted March 25, 2013 I went from 3.1 to 3.3 with no changes to downloading or habit. First 4gb file I download and I got that dreaded "Disk Overload". Never a problem before. Turned off the partial settings, never do partial downloads anyway, tried with cache on/off and no change, upped cache to 1.5gb and was ok for a bit but then the error. Would be crazy to keep a cache that large anyways. I rolled back to 3.2.28705 and my 4gb file went without a problem. I also see some torrent sites are starting to Ban uTorrent 3.3 as they dont want to deal with the the problem and may ban it on my site also.By the way, my drives are a Sata Raid0 setup. And as I said I have had no problems prior to 3.3 and also have kept a 256mb cache. Cant believe this has been going on this long and not internally tested. Link to comment Share on other sites More sharing options...
rafi Posted March 25, 2013 Report Share Posted March 25, 2013 I went from 3.1 to 3.3 Try 3.3.1 + my settings.dat file... Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.