[freenet-dev] Re: Profiling of fred

2003-05-21 Thread Toad
On Wed, May 21, 2003 at 01:39:41PM -0400, Dan Merillat wrote: > On Mon, 19 May 2003, Toad wrote: > > > > > Because we do not have access to the lastaccess time in java. And also > > because we need to keep the in-memory LRU consistent and up to date. > > Then I humbly suggest only adjusting it o

[freenet-dev] Re: Profiling of fred

2003-05-21 Thread Dan Merillat
On Wed, 21 May 2003, Toad wrote: > On Wed, May 21, 2003 at 01:39:41PM -0400, Dan Merillat wrote: > > On Mon, 19 May 2003, Toad wrote: > > > > > > > > Because we do not have access to the lastaccess time in java. And also > > > because we need to keep the in-memory LRU consistent and up to date.

[freenet-dev] Re: Profiling of fred

2003-05-21 Thread Dan Merillat
On Mon, 19 May 2003, Toad wrote: > > Because we do not have access to the lastaccess time in java. And also > because we need to keep the in-memory LRU consistent and up to date. Then I humbly suggest only adjusting it on file open and close rather then every read. This also saves us a ton of L

[freenet-dev] Re: Profiling of fred

2003-05-20 Thread Niklas Bergh
>>NativeFSDir level because it might interfere with CircularBuffer >>implementation. In CVS unstable branch, please update and try again. > >Good, I will do a shot on it as soon a possible to see if I can provide any hard numbers on the exact performance gain. No hard numbers but a couple of scre

[freenet-dev] Re: Profiling of fred

2003-05-19 Thread Toad
On Mon, May 19, 2003 at 03:52:39PM -0400, Dan Merillat wrote: > On Sat, 17 May 2003, Toad wrote: > > > On Sat, May 17, 2003 at 02:05:44PM +0200, Niklas Bergh wrote: > > > Well, I guess I just have to start the discussion myself then :).. > > > > > > According to the profiler 30% of the CPU is spe

[freenet-dev] Re: Profiling of fred

2003-05-19 Thread Dan Merillat
On Sat, 17 May 2003, Toad wrote: > On Sat, May 17, 2003 at 02:05:44PM +0200, Niklas Bergh wrote: > > Well, I guess I just have to start the discussion myself then :).. > > > > According to the profiler 30% of the CPU is spent in > > NativeFSDirectory$NativeBuffer.setLastModified.. Well.. this met

[freenet-dev] Re: Profiling of fred

2003-05-18 Thread Niklas Bergh
>>On Sat, May 17, 2003 at 02:05:44PM +0200, Niklas Bergh wrote: >> Well, I guess I just have to start the discussion myself then :).. >> >> According to the profiler 30% of the CPU is spent in >> NativeFSDirectory$NativeBuffer.setLastModified.. Well.. this method would >Hrmm. This should usually b

[freenet-dev] Re: Profiling of fred

2003-05-18 Thread Niklas Bergh
>> of data that I was requesting from the node (2.5 megs). See the attached >> HTML if you want to study an example of this situation yourself (just expand >> the tree until you encounder the invalid percentages and then continue >> expanding). >Okay, the problem seems to be single byte reads from

[freenet-dev] Re: Profiling of fred

2003-05-18 Thread Niklas Bergh
>> Setup: >> 1. Firewalled v6032 node (not transient but not able to recieve >> incoming connections) >If it cannot receive incoming connections it should be set transient. True, it was an accident which I discovered halfway through the profiling. This might be a good time to start a discussion

[freenet-dev] Re: Profiling of fred

2003-05-17 Thread Toad
On Fri, May 16, 2003 at 07:31:33PM +0200, Niklas Bergh wrote: > Damn, I sent this mail to myself instead of to devl.. Another try then.. > sorry for loosing the formatting. > > The attachements I am talking about can be retrieved from > http://194.236.28.174/freenetstuff/prof.zip > > /N > >

[freenet-dev] Re: Profiling of fred

2003-05-17 Thread Toad
On Sat, May 17, 2003 at 02:05:44PM +0200, Niklas Bergh wrote: > Well, I guess I just have to start the discussion myself then :).. > > According to the profiler 30% of the CPU is spent in > NativeFSDirectory$NativeBuffer.setLastModified.. Well.. this method would > seem to be called close to once

[freenet-dev] Re: Profiling of fred

2003-05-17 Thread Toad
On Sat, May 17, 2003 at 02:05:44PM +0200, Niklas Bergh wrote: > Well, I guess I just have to start the discussion myself then :).. > > According to the profiler 30% of the CPU is spent in > NativeFSDirectory$NativeBuffer.setLastModified.. Well.. this method would Hrmm. This should usually be pret

[freenet-dev] Re: Profiling of fred

2003-05-17 Thread Niklas Bergh
Well, I guess I just have to start the discussion myself then :).. According to the profiler 30% of the CPU is spent in NativeFSDirectory$NativeBuffer.setLastModified.. Well.. this method would seem to be called close to once for every single byte read from the datastorage. That method is called 2

[freenet-dev] Re: Profiling of fred

2003-05-16 Thread Niklas Bergh
Damn, I sent this mail to myself instead of to devl.. Another try then.. sorry for loosing the formatting. The attachements I am talking about can be retrieved from http://194.236.28.174/freenetstuff/prof.zip /N - Original Message - From: "Niklas Bergh" To: Sent: Friday, May 16, 2003 5