CPU % increase

[YoThisBling]

CPU % increase
4.11.1 and 4.11.2 betas, the CPU% increases much more than previous releases.

CPU% spikes to 90% for a long period of time.
What is the casue of this?

[stief]

I'm not sure of the cause, because I'm on a Mac and have had Apple and Java problems for a while.

on a G4 running 10.4.6 and Java 1.5, I've noticed CPU as much higher than before. 25% used to be expected, but with the betas it is up to more than 40%, and frequently at 70%.

Can't say I've noticed any steady 90% though. How much RAM do you have?

[Sam]

Yes, the CPU will be slightly higher with the newer versions. Older versions had an artificial limitation on the bandwidth, which kept the CPU lower but reduced the overall network bandwidth. We've removed this limitation, so downloads and uploads should become much faster as people upgrade, but the CPU will be a little bit more taxed.

[stief]

a "little bit more taxed" I expected. But should double or more CPU use be expected? I really haven't changed my upload/download habits significantly . . .

Can the LW tests give an approximate % of what increase we should expect?

[YoThisBling]

I'm not sure of the cause, because I'm on a Mac and have had Apple and Java problems for a while.

on a G4 running 10.4.6 and Java 1.5, I've noticed CPU as much higher than before. 25% used to be expected, but with the betas it is up to more than 40%, and frequently at 70%.

Can't say I've noticed any steady 90% though. How much RAM do you have?
I have 1 GB of RAM and it has stayed around 90%.

[Agent_J]

LimeWire 4.13.0 Beta still seems to have this problem, any chance these suggestions could be (re)implemented?

http://www.limewire.org/techdocs/performance.htm

[Sam]

Sure.

Add support for non-blocking IO, which greatly reduces the number of threads.
Mostly finished a few versions ago, 99% finished with 4.12.

Reduce startup time by delaying the loading of inessential GUI components, like the options window.
Finished many versions ago.

Optimize connection initialization/handshaking. There appears to be a number of reasons for poor performance. First, the calls to HostCatcher.add can be too slow, which is called for every X-Try/X-Try-Ultrapeer address in connection handshake headers. Second, the act of actually sending and receiving headers (in AuthenticationHandshakeResponder and subclasses) can be slow.
Finished many versions ago. The whole Authentication series of classes removed, HostCatcher.add takes a trivial amount of time to run, and X-Try headers are very rarely used.

Reduce memory footprint of query routing (QRP) tables. Currently we allocate a whole byte per entry; we only need a bit.
Finished many, many versions ago.

A better approach might be to use a single route table for all connections. Each entry of this table would be a list of all connections matching that index,
Finished many, many versions ago.

Optimize message writing in ManagedConnection. Writing a message involves queuing a message and notifying its associated output thread. The output thread then removes the message from the queue and writes it to the socket. There are two things to optimize: the overhead of wait/notify and of manipulating the queue. The latter is easier to optimize.
Finished with the introduction of NIO. The thread overhead is obliterated because there's no more threads, and queue performance is vastly increased.

Also, it should be possible to reduce the memory footprint of these queues. Leaf connections may not need the full SACHRIFC buffers, for example. It is probably not necessary to preallocate a fixed amount of memory for these buffers when they are typically mostly empty.
Leaves still have SACHRIFC-type buffers set up, but memory is lazily allocated for them.

At the least, we should avoid allocating these buffers until the connection has been initialized and accepted for normal use.
Done with the introduction of NIO (although it's unrelated).

Reduce the memory footprint needed to display search results.
Done many versions ago.

Optimize ConnectionManager.hasClientSupernodeConnection. Currently this method can use up to 20% of all CPU time. The problem is that it iterates through all connections, which requires cloning the array. This method is called from allowAnyConnection, which is called when accepting new connections or handling pings. A better approach is to simply to augment ConnectionManager with a boolean that is true iff one of the connection is a leaf to ultrapeer connection. Be sure to maintain the invariant when adding and removing connections.
Done many versions ago.

Optimize the creation of GGEP extensions in pongs. Currently it can consume 20% of all message handling time. The problem is too many calls to StringBuffer.append from PingReply.write, which is called from PingReply.newGGEP from the PingReply constructor.
So very done.

Optimize route table GUID lookup. Currently we use TreeMap to back our RouteTable data structure. I can't remember why we don't use HashTable's; I think it has to do with performance problems from resizing the tables. I would be curious to try splay trees, a randomized data structure. Splay trees can outperform standard dictionaries under irregular access patterns, e.g., when handling a stream of query replies with the same GUID..
Not done, but not really a problem still.

We should probably update all those documents. Pretty much every one is grossly out-of-date.

[stief]

:p

Thanks for taking the time to summarize all those improvements. . . amazing how much has been done, especially with NIO.

Still, twice the CPU has stuck way high (+90%) on a Linux system (after running more than a couple of days continuously) . Initially the CPU usage, even as an UP, was very low--6-10%

I figured it was my inexperienced setup in Linux, so I wiped the drive and reinstalled from scratch. Then it happened again.

Quick tips for capturing this on Linux much appreciated.

[vip8439]

Mine usually spikes at %100 and I cant switch between opened windows without waiting for 5 minutes. I have a 1.8 Ghz CPU and 512MB of RAM. This happens in XP, Server 2003, and Fedora Core. I'm kinda surprised to see it is a removed artificial bandwith limitation that is the culprit. In a post I've read before, LimeWire said it was a Java bug. That may be so, but its kind of funny how none of my other Java applications bring my computer to a halt. Also funny how I can run Doom3 at a very acceptable rate, but LimeWire brings my comp to its knees. What are the developers using? Quad, Intel Core 2 Duo machines for testing? Whatever it is, I wish LimeWire would fix it. I'm very tired of this and really don't care for many of the P2P alternatives except for Ares.

[Sam]

Very very few other Java programs have as intense a network stack as LimeWire. We kind of use Java to the extreme.

[snoopy]

Are they working on this CPU problem?

This is unaccptable to eat up 95-100% CPU.

Temp fix, exit LW & restart to reduce CPU use to ~20%

I tried the suggestion of removing incomplete download files (useless since "Could not continue Previous downloads" message (can they fix that too)

SERIOUS PROBLEMS that need to be fixed for next release!
- high CPU usage
- not responding after latest Java installed
- frequent crashing & unable to continue previous downloads (continue off files not data files?)

[lembas]

Yes, the CPU will be slightly higher with the newer versions. Older versions had an artificial limitation on the bandwidth, which kept the CPU lower but reduced the overall network bandwidth. We've removed this limitation, so downloads and uploads should become much faster as people upgrade, but the CPU will be a little bit more taxed.

What is the latest version having the artificial cpu limitation?

[snoopy]

I'm using the latest Beta version. I had the same problem with the latest non-beta version.

Sorry, I noticed beta continues downloads are it frequently locks up & I have to faorce an end task thru Win XP.

I tried different java versions & all same problem, LW hits 99% cpu or doesn't respond (not necissarily both)

I'm seeing posts others are having these serious problems & they really need to be fixed before beta is released.

Is there a debug mode to track down the problem?

There must be something that can be done, else it's hopeless & we'll all have to stop using it :(

Should we give up & use FrostWire.com?

[snoopy]

Well, I haven't heard of anything on this & since by the vote 77% of the people are having the problem I was wondering ... is this being looked into?

There clearly seems a problem since with the same bandwidth the CPU spikes from ~25% to 95-100% then if it at least responds so you can exit it then restart it it goes back to the ~25% while still using the same bandwidth.

Should we give up on LimeWire & just goto FrostWire?

[Sam]

I doubt FrostWire will be any different, being as it's based 100% on LimeWire's code. You're welcome to try, though.

[Lord of the Rings]

Yep I've seen reports of it happening to FrostWire also. But it's only affecting some users of both programs. Probably a small % overall. Otherwise GF would be inundated with such complaints.

[lembas]

version 4.10 is as fast as the last version for me. use 4.10. you can dind it on outofdatefadversion.com

exact link: to here (http://www.limewire.org/forum/showthread.php?t=125&page=2)

i tried frostwire. it has the problem. do not use it.

[snoopy]

You mentioned LW 4.10 is just as fast as latest version, but that says nothing of the CPU problem being discussed ... does 4.10 have that same problem? Ugh! Seems noones looking into this & the problem will be replicated to Frostwire since it's not being fixed or looked into :(

[Sam]

It's being looked into.

[Grandpa1]

Just curious as to whether any of you have tried Java 6 and seen if that helped with the problem.

http://java.sun.com/javase/downloads/index.jsp

[chris999]

It seems to me that as soon as an upload starts the cpu load on my machine jumps to 100%. Since my link is limited to 2k per second this obviously isn't just normal java load it seems more like a bug in the upload code, and judging from this forums it's a new bug introduced when the limits were removed on upload speed..., netscape used to have a similar bug, probably a loop in the code trying to send data when the socket write call is failing with 'try later'

Anyway that's my guess. Appologies if I"m wrong as wrong can be :-)

I'm using version 4.12.6, java 1.5.0_06

Any advice appreciated.

[Sam]

Recognizing it's when uploads happen is very useful. More insights like this would be great in tracking down the increase.

[Xcom46]

In some cases i have had with my Cpu runing at 100% i found a good fix Try to air hos your pc out sometimes it might get stuck runing at 90 to 100% because the fan is turning to fast and it might be just burning up so try to air hos it out once i a while like every 6 months or so.after i did mine it worked fine.


Chris B.......

[zab]

It seems to me that as soon as an upload starts the cpu load on my machine jumps to 100%. Since my link is limited to 2k per second this obviously isn't just normal java load it seems more like a bug in the upload code, and judging from this forums it's a new bug introduced when the limits were removed on upload speed..., netscape used to have a similar bug, probably a loop in the code trying to send data when the socket write call is failing with 'try later'

Anyway that's my guess. Appologies if I"m wrong as wrong can be :-)

I'm using version 4.12.6, java 1.5.0_06

Any advice appreciated.


Can you see if this happens with 4.13?

[zab]

Also, whenever you get a 100% cpu usage few seconds log of the com.limegroup.gnutella.io.* package would be very useful. There should be instructions around here how to get such a log.

[JohnTheNutter]

Obviously everyone is aware that the beta 6 Java is no longer a beta? JRE 6 is available here http://java.sun.com/javase/downloads/index.jsp (well its updated this thread anyway!!)

[lembas]

Frostwire has the same problem. do not use it. the only solution is limewire version 4.10. oldversion.com has this version.

[JohnTheNutter]

This does not seem to be a problem on the 4.13 Beta?? Or am I just lucky?

[snoopy]

Yep I've seen reports of it happening to FrostWire also. But it's only affecting some users of both programs. Probably a small % overall. Otherwise GF would be inundated with such complaints.

It's not good to assume, often people don't take the time to complain because it takes time, maynot help or because someone else reported the problem (hint, hint) plus the voting on this showed many ARE having the problem.

Also PLEASE read previous posts. Many asked if this is happening in the 4.13 beta version & I mentioned in the begining that it is.

It was also mentioned that it's being looked into (any progress?). I hope it gets fixed before 4.13 is released. It seems silly to have frostwire & limewire running side by side in the first place let alone with such serious bugs ... why not just have one vs 2 programs to maintain?

Certainly this can be traced & debugged to find the problem

[Sam]

We don't maintain FrostWire, we have nothing to do with it. It's other people that are using LimeWire's code and rebranding it.

[_Sara_]

I am running 4.13 BETA Pro. I have tried everything mentioned here. I have nothing being uploaded, I am running an AMD 64bit 3400+ with 1gig + 256mb of RAM and I have updated my Java, yet after a couple of minutes my Limewire always maxes out my CPU to 100%. I have tried closing the search windows, stopping searches early so that its not constantly looking, I have tried closing and restarting so that only the downloads are loaded, and I have cleared out ALL stalled downloads there are only 3 in my que and all are active but it still goes to 100% after a few minutes.

[snoopy]

Any progress on the high CPU problem?

Any debugging software like strobe for the mainframes that can tell you what lines are eating up the CPU?

It seems to be happening far more often lately.

As you can see here, MANY others are having the problems also.

http://www.google.com/search?hl=en&q=limewire+problem+CPU

Anyone suggest alternate software that works well?

[Xcom46]

You can Try
System Mechanic 7

Finds all kinda porblems I got mine at wallmart for 20$
Than Check this program out.It's called System Mechanic 7 Works Very well help's with slowing pc's crashing pc's and bad internet speed's for like Dsl and other it fix min up very good and i did not no that i had all these problems with mine it was worth the 20.00$ You can get it at walmart well that's where i got mine any way.

Let me Explain how it works.
Fixes critical Problems
Warnings
Your Memory level
Registry
-----------------------
Increase Performance.
Pc Accelerator.
Defragment Hard Drives.
OptimizezWindows Start Up.
Optimizez Internet Settings.
Defragment Memory Or it's Called Ram.
Compact Registry Files.
Configure Boot-Time Defragmentaion.
Enhance Protection.
Free up Hard Drive Space.
Repair Problems.
Ensure Personal Privacy.
Manage System Configurations.
Perform Diagnostics.

Works well Go out and get it like i said worth 20.00$ if your pc is runing badly.

Chris B.....

[snoopy]

From your description, it seems it can't indicate as mentioned the LimeWire code that's eating up CPU ... so it seem it wouldn't help for this situation.

Let me know if I'm wrong & if it can do that. Thanks for the post.

If anyone know software to do this please let us know. Strobe Mainframe software does this don't know if there's anything for PC's like that

[Xcom46]

A lot of time's it's registry that use's a lot of cpu you might havea lot registry problems and not even no about that all slows down pc and it do's help find broken link's in program's and it will fix tham.it do's a deep scan for about 10 min's and all so depends on how big your system is.it fixed a lot of problems on my computer and i did not even no about them works well.

I found a link at download.com see if it works for you.
http://www.download.com/System-Mechanic-Professional/3000-2094_4-10659866.html?tag=lst-0-2

Chris B.......

[snoopy]

Thanks but System Mechanic doesn't seem it will help.

Also I'm seeing problems online re: System Mechanic & it being classify as malware. http://www.wrensoft.com/forum/showthread.php?t=1451

It's a known problem that LimeWire has code that's eating up CPU & putting it at nearly 100% since it happens OFTEN in LimeWire & only when it's running.

To avoid clutter on this post, please reply to only me in person regarding this further.

[Only A Hobo]

@ snoopy, we really don't mind cluttered threads that much, especially if they contain information that might be helpful or interesting for others, and this one does seem to :) :)

[derekmski]

CPU % increase
4.11.1 and 4.11.2 betas, the CPU% increases much more than previous releases.

CPU% spikes to 90% for a long period of time.
What is the casue of this?


Yes, how much RAM and processor power does your computer have? I have 3.2GHz Intel pentium D, 2GB of RAM, Windows Vista Home Premium (32 bit) and I have no problems with my CPU % spiking. It spikes to 100% while loading then drops down to 5-15%

[derekmski]

I have like I said a 3.2GHz Pentium D 960, 2GB of RAM, Windows Vista Home Premium (32-bit) and I am seeing no spike except for the loading of limewire but then it goes way down to nothing? Maybe your guys computers can't handle it? It is really weird I see absolutely no spike except when loading like i said then it works great after that.

[Xcom46]

I have a 1.40Ghz Amd
With 3Gb of Ram
G force fx 5200 video card
C: Drive 40Gig
E: Drive 160Gig
CDRom Dvd-r-w-burner

Chris B....

[Xcom46]

I only have a problem when i dont defrag my drive's four about three days.

[snoopy]

People, please read the past posts. There's lots of info being duplicate cause people are not reading them.

I have 512 M RAM, but others are having the problem with 1.25 Gig & 3 gig of RAM besides RAM would have nothing to do with CPU, that's I/O & I'm not even paging that much so that's clearly not the issue.

You're never having your LimeWire CPU spike at all ever on Vista? I find that hard to believe, mine's happening every couple hours of running & I have to stop & restart it. Maybe somethings done different or there's a bug that XP is catching with LW.

Sorry to be harsh it's just frustrating this seems to have been a big problems for a long time & it seems no progress is being made & talking of all this it seems it will help little vs what's really needed .... debugging & finding the code that's eating up the CPU & fixing it.

Anyone doing that?

Any progress?

[derekmski]

People, please read the past posts. There's lots of info being duplicate cause people are not reading them.

I have 512 M RAM, but others are having the problem with 1.25 Gig & 3 gig of RAM besides RAM would have nothing to do with CPU, that's I/O & I'm not even paging that much so that's clearly not the issue.

You're never having your LimeWire CPU spike at all ever on Vista? I find that hard to believe, mine's happening every couple hours of running & I have to stop & restart it. Maybe somethings done different or there's a bug that XP is catching with LW.

Sorry to be harsh it's just frustrating this seems to have been a big problems for a long time & it seems no progress is being made & talking of all this it seems it will help little vs what's really needed .... debugging & finding the code that's eating up the CPU & fixing it.

Anyone doing that?

Any progress?

Yes Limewire 4.13 beta works great on windows vista, I don't have it run steady I have it shutdown imediatley when done because I am a gamer and I need all the RAM but when it runs and I download 5 or so songs at a time she works fine.

[snoopy]

Thanks, but downloading 5 songs is not a good test

It usually doens't spike to ~99% CPU till ~2-3 hours use. Thanks anyway

[derekmski]

Thanks, but downloading 5 songs is not a good test

It usually doesn't spike to ~99% CPU till ~2-3 hours use. Thanks anyway

Well then shame on you for using it 24/7. The record companies will catch you if you leave Limewire on steady. You download a song, you leave Limewire open, they track you, and fine you a couple million dollars for download so, your choice?

[Aaron.Walkhouse]

FUD! That's not how they find people.

If you don't share their stuff, they can't see you at all, period, end of sentence.

The best way to use any gnutella servent is to keep it
running permanently, which guarantees your shares can be
found and enables you to search for and download the
rarest files.

[derekmski]

FUD! That's not how they find people.

If you don't share their stuff, they can't see you at all, period, end of sentence.

The best way to use any gnutella servent is to keep it
running permanently, which guarantees your shares can be
found and enables you to search for and download the
rarest files.

Well sucks to be you because I am having zero problems so ha ha ha sucker.

[Xcom46]

Yea there R many ways to get your self busted on the net even if your really carfull but there R all so many ways not to get busted but all in all if they whant to sue.Than there going to have to sue every one to.It would be a bad for the people that sue other people.every thing would drop in price sales would be low for them that's Y they dont sue every one.
It's most likely the Ip that are talking to the feds or what ever other goverment there is.


Chris B......

[Lord of the Rings]

I have no problems with my CPU % spikingNot too sure why this person even bothered to post in this thread. Hasn't said much if anything constructive. But has been insulting. And from what I hear, I definitely would not recommend a Dell. Too bad if you need to reformat your entire system heh. lol :D Sounds like someone not interested in sharing. Let's all not share for a day & see how much he is able to download. :p

[Xcom46]

I argee with you LOTR
Probley has a dell and dell sucks ***
any thing with A Amd i will take but i will not take a dell or Hp they are so slow even with a 2GHZ proceser i have a hp at work with 2ghz intel it take 30seconds for stuff to pop up.even if the hard drive is defraged.and with mine it take 2 seconds for every thing to pop up and im runing a 1.40Amd Ghz with 3gb of ram two hard drive's 40g&160.And all so DvD-RW Drive 52X A G Force Fx 5200 Video card Mine Rocks.

[AngilTDS]

My CPU maxes out within minutes of starting LimeWire, and my machine does not max out easily. The one trend that I have noticed, is that it seems to do it after it has cycled through the first 5000 ports, and starts back over. You might want to make sure that it is recycling the ports correctly, and that there are conflicts between what xp allows on the ports, and what it is using them for. Once it reaches 100% cpu use, it will stay there even if I clear out all downloads, and there are no uploads. Only way to free cpu is to shut LW down.

Hope this Helps.

Angil
AMD 64 FX-5
2Gig Corsair
Raided Raptors
2+tb
Xp with Latest Java & LW