[freenet-dev] Garbage collection /Performance?

2006-07-11 Thread Matthew Toseland
ible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/ffeb784f/attachment.pgp>

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Florent Daignière (NextGen$)
* Jusa Saari [2006-07-11 21:58:26]: > On Sun, 09 Jul 2006 10:40:17 +0200, Florent Daigni?re wrote: > > > * Colin Davis [2006-07-08 23:41:24]: > > > >> Fair- The profilers I found all seemed to be GUI based- I run my node on > >> a server without X, that I can only access via SSH. I know Java

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Matthew Toseland
: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/cf69a6b3/attachment.pgp>

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Jusa Saari
On Sun, 09 Jul 2006 10:40:17 +0200, Florent Daigni?re wrote: > * Colin Davis [2006-07-08 23:41:24]: > >> Fair- The profilers I found all seemed to be GUI based- I run my node on >> a server without X, that I can only access via SSH. I know Java 1.5 has >> a way do generate a profile text file,

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Jusa Saari
Or just use the -XX:+DisableExplicitGC parameter to make Sun JVM ignore garbage collection requests. Simply add wrapper.java.additional.1=-XX:+DisableExplicitGC to wrapper.conf, like I just did. On Mon, 10 Jul 2006 22:54:07 +0100, Matthew Toseland wrote: > I would support making it condition

[freenet-dev] Re: [freenet-cvs] r9529 - trunk/freenet/src/freenet/node

2006-07-11 Thread Florent Daignière (NextGen$)
* Matthew Toseland [2006-07-11 00:52:11]: > Ummm... > > Please don't take the lock on run() This is a long method which > does lots of blocking I/O ! I wouldn't be surprised if this patch causes > deadlocks... Also iirc finish() can do I/O, so again it shouldn't be > locked. htl will not be

[freenet-dev] Re: Generating more freenet content

2006-07-11 Thread Matthew Toseland
Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/32bfd09b/attachment.pgp>

[freenet-dev] Re: Generating more freenet content

2006-07-11 Thread Yongqian Li
l the things you care about in one place. http://www.live.com/getstarted -- next part -- An HTML attachment was scrubbed... URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/af6a692c/attachment.html>

[freenet-dev] Garbage collection /Performance?

2006-07-11 Thread David Sowder (Zothar)
Matthew Toseland wrote: > I would support making it conditional on a flag; node.aggressiveGC=<-1 > or period in seconds>. > > On Sun, Jul 09, 2006 at 11:17:13AM +0200, Florent Daigni?re (NextGen$) wrote: > >> * Ian Clarke [2006-07-09 02:05:43]: >> >> >>> On 9 Jul 2006, at 01:40, Florent Da

[freenet-dev] Re: [freenet-cvs] r9569 - trunk/freenet/src/freenet/node

2006-07-11 Thread Matthew Toseland
Expand the synchronization on the if(cw == null) { set cw }; if you check a variable and then set it, the check and the set should be synchronized (so it is atomic). On Tue, Jul 11, 2006 at 10:58:27PM +, [EMAIL PROTECTED] wrote: > Author: nextgens > Date: 2006-07-11 22:58:25 + (Tue, 11 Jul

Re: [freenet-dev] Garbage collection /Performance?

2006-07-11 Thread Matthew Toseland
On Tue, Jul 11, 2006 at 05:40:22PM -0500, David Sowder (Zothar) wrote: > >> > On my node, which is admittedly not setup to use more than the default > 128 MB of RAM, the MemoryChecker run GC is done roughly every 30 seconds > and with the timing and before/after compare I've added locally, it

Re: [freenet-dev] Garbage collection /Performance?

2006-07-11 Thread David Sowder (Zothar)
Matthew Toseland wrote: I would support making it conditional on a flag; node.aggressiveGC=<-1 or period in seconds>. On Sun, Jul 09, 2006 at 11:17:13AM +0200, Florent Daignière (NextGen$) wrote: * Ian Clarke <[EMAIL PROTECTED]> [2006-07-09 02:05:43]: On 9 Jul 2006, at 01:40, Florent

Re: [freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Matthew Toseland
On Tue, Jul 11, 2006 at 11:27:46PM +0200, Florent Daignière (NextGen$) wrote: > > > > What did I miss ? > > That the base object in freenet is the "Bucket" and that we do use > finalizers around it :) Buckets should be explicitly free()d. If we don't, there's a space leak on disk. > > NextGen$

Re: [freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread NextGen$
* Jusa Saari <[EMAIL PROTECTED]> [2006-07-11 21:58:26]: > On Sun, 09 Jul 2006 10:40:17 +0200, Florent Daignière wrote: > > > * Colin Davis <[EMAIL PROTECTED]> [2006-07-08 23:41:24]: > > > >> Fair- The profilers I found all seemed to be GUI based- I run my node on > >> a server without X, that I

[freenet-dev] Re: [freenet-cvs] r9533 - in trunk/freenet/src/freenet: clients/http/filter support

2006-07-11 Thread Matthew Toseland
/freenet/support/UpdatableSortedLinkedList.java > >>=== > >>--- trunk/freenet/src/freenet/support/UpdatableSortedLinkedList.java > >>2006-07-09 22:47:52 UTC (rev 9532) > >>+++ trunk/freenet/src/freenet/support/UpdatableSortedLinkedList.java > >>2006-07-09 23:21:53 UTC (rev 9533) > >>@@ -54,13 +54,10 @@ > >> throw new IllegalStateException("impossible"); > >> } > >> > >>-private StringBuffer sb = new StringBuffer(1000); > >>- > >> protected synchronized void checkList() { > >>// FIXME once satisfied that this works, make it only happen > >>occasionally > >>int statedLength = list.size(); > >>int realLength = 0; > >>- sb.setLength(0); > >>for(Enumeration e = list.elements();e.hasMoreElements();) { > >>UpdatableSortedLinkedListItem i = > >>(UpdatableSortedLinkedListItem) e.nextElement(); > >>// Sanity check for infinite looping > >> > >>___ > >>cvs mailing list > >>cvs at freenetproject.org > >>http://emu.freenetproject.org/cgi-bin/mailman/listinfo/cvs > >> > >> > > > > > > > > > >___ > >cvs mailing list > >cvs at freenetproject.org > >http://emu.freenetproject.org/cgi-bin/mailman/listinfo/cvs > > ___ > Devl mailing list > Devl at freenetproject.org > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl > -- Matthew J Toseland - toad at amphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/4ab433dd/attachment.pgp>

[freenet-dev] Re: [freenet-cvs] r9540 - in trunk/freenet/src/freenet: client client/async support

2006-07-11 Thread Matthew Toseland
ctx, this, isMetadata, > sourceLength, token, getCHKOnly, false, true /* we don't use it */, > tokenObject); > - } catch (InserterException e) { > - cb.onFailure(e, this); > - return; > - } > + if(finished) return; > + edition = edNo; > + Logger.minor(this, "scheduling insert for "+pubUSK.getURI()+" > "+edition); > + try { > + sbi = new SingleBlockInserter(parent, data, > compressionCodec, privUSK.getInsertableSSK(edition).getInsertURI(), > + ctx, this, isMetadata, sourceLength, > token, getCHKOnly, false, true /* we don't use it */, tokenObject); > + } catch (InserterException e) { > + cb.onFailure(e, this); > + return; > } > try { > sbi.schedule(); > @@ -122,13 +117,11 @@ > } > } > > - public void onSuccess(ClientPutState state) { > + public synchronized void onSuccess(ClientPutState state) { > cb.onEncode(pubUSK.copy(edition), this); > cb.onSuccess(this); > - synchronized(this) { > - finished = true; > - sbi = null; > - } > + finished = true; > + sbi = null; > FreenetURI targetURI = pubUSK.getSSK(edition).getURI(); > FreenetURI realURI = ((SingleBlockInserter)state).getURI(); > if(!targetURI.equals(realURI)) > @@ -140,7 +133,7 @@ > // FINISHED Yay!!! > } > > - public void onFailure(InserterException e, ClientPutState state) { > + public synchronized void onFailure(InserterException e, ClientPutState > state) { > sbi = null; > if(e.getMode() == InserterException.COLLISION) { > // Try the next slot > @@ -195,7 +188,7 @@ > scheduleInsert(); > } > > - public void onCancelled() { > + public synchronized void onCancelled() { > if(finished) return; > Logger.error(this, "Unexpected onCancelled()", new > Exception("error")); > cancel(); > > Modified: > trunk/freenet/src/freenet/support/SimpleIntNumberedItemComparator.java > === > --- trunk/freenet/src/freenet/support/SimpleIntNumberedItemComparator.java > 2006-07-10 01:50:31 UTC (rev 9539) > +++ trunk/freenet/src/freenet/support/SimpleIntNumberedItemComparator.java > 2006-07-10 10:22:11 UTC (rev 9540) > @@ -12,7 +12,7 @@ > > public int compare(Object o1, Object o2) { > int x = ocompare(o1, o2); > -Logger.minor(this, "compare("+o1+","+o2+") = "+x); > + Logger.debug(this, "compare("+o1+","+o2+") = "+x); > return x; > } > > > ___ > cvs mailing list > cvs at freenetproject.org > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/cvs > -- Matthew J Toseland - toad at amphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/9bca4670/attachment.pgp>

[freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
nd - toad at amphibian.dyndns.org > Freenet Project Official Codemonkey - http://freenetproject.org/ > ICTHUS - Nothing is impossible. Our Boss says so. > ___ > Devl mailing list > Devl at freenetproject.org > http://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl -- Matthew J Toseland - toad at amphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/a3eb21cf/attachment.pgp>

[freenet-dev] Opennet warning?

2006-07-11 Thread jdavie...@comcast.net
achment was scrubbed... URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/06703fb1/attachment.html>

[freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/7c0bf0ea/attachment.pgp>

[freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
mphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/2ba21022/attachment.pgp>

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Jusa Saari
On Sun, 09 Jul 2006 10:40:17 +0200, Florent Daignière wrote: > * Colin Davis <[EMAIL PROTECTED]> [2006-07-08 23:41:24]: > >> Fair- The profilers I found all seemed to be GUI based- I run my node on >> a server without X, that I can only access via SSH. I know Java 1.5 has >> a way do generate a

[freenet-dev] Re: Garbage collection /Performance?

2006-07-11 Thread Jusa Saari
Or just use the -XX:+DisableExplicitGC parameter to make Sun JVM ignore garbage collection requests. Simply add wrapper.java.additional.1=-XX:+DisableExplicitGC to wrapper.conf, like I just did. On Mon, 10 Jul 2006 22:54:07 +0100, Matthew Toseland wrote: > I would support making it condition

Re: [freenet-dev] Re: Generating more freenet content

2006-07-11 Thread Matthew Toseland
It's encrypted, but the encryption key (for CHKs, which are used for bulk data storage) is derived from the original data, so it is deterministic. For SSKs it is random, but SSKs are only used for freesite redirects and so on; they're only 1kB of payload. On Tue, Jul 11, 2006 at 06:17:19PM +,

[freenet-dev] Re: Generating more freenet content

2006-07-11 Thread Yongqian Li
>>As I understand it, there isn't harm in inserting the same page >>multiple times, since it will just collide on the CHK, and not be >>stored twice.I'm not sure but I don't think it will collide since the content is encrypted before being inserted. Since the encryption key is random, the cont

Re: [freenet-dev] Re: [freenet-cvs] r9529 - trunk/freenet/src/freenet/node

2006-07-11 Thread NextGen$
* Matthew Toseland <[EMAIL PROTECTED]> [2006-07-11 00:52:11]: > Ummm... > > Please don't take the lock on run() This is a long method which > does lots of blocking I/O ! I wouldn't be surprised if this patch causes > deadlocks... Also iirc finish() can do I/O, so again it shouldn't be > lock

[freenet-dev] Opennet warning?

2006-07-11 Thread Florent Daignière (NextGen$)
the default behaviour. NextGen$ -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/9d92227f/attachment.pgp>

[freenet-dev] Re: [freenet-cvs] r9533 - in trunk/freenet/src/freenet: clients/http/filter support

2006-07-11 Thread David Sowder (Zothar)
Matthew Toseland wrote: > Can't we just clear() it? What does lint4j have against > StringBuffer.setLength() exactly? > Because I didn't see nor do I see a clear() method in the docs for StringBuffer nor did lint4j suggest it and I'm still a relative newb to the Java API? lint4j says it's a pe

Re: [freenet-dev] Re: [freenet-cvs] r9533 - in trunk/freenet/src/freenet: clients/http/filter support

2006-07-11 Thread Matthew Toseland
On Tue, Jul 11, 2006 at 08:04:12AM -0500, David Sowder (Zothar) wrote: > Matthew Toseland wrote: > >Can't we just clear() it? What does lint4j have against > >StringBuffer.setLength() exactly? > > > Because I didn't see nor do I see a clear() method in the docs for > StringBuffer nor did lint4j

Re: [freenet-dev] Opennet warning?

2006-07-11 Thread jdaviestx
> Hmmm, on second thoughts, lets not take this metaphor too far... if> opennet mode is promiscuous, then darknet is polyamorous...Good point - If my dates were in promiscuous mode by default, I doubt I'd switch them over... ___ Devl mailing list Devl@fr

[freenet-dev] Re: [freenet-cvs] r9533 - in trunk/freenet/src/freenet: clients/http/filter support

2006-07-11 Thread David Sowder (Zothar)
Matthew Toseland wrote: Can't we just clear() it? What does lint4j have against StringBuffer.setLength() exactly? Because I didn't see nor do I see a clear() method in the docs for StringBuffer nor did lint4j suggest it and I'm still a relative newb to the Java API? lint4j says it's a perfor

[freenet-dev] Re: [freenet-cvs] r9540 - in trunk/freenet/src/freenet: client client/async support

2006-07-11 Thread Matthew Toseland
On Mon, Jul 10, 2006 at 10:22:21AM +, [EMAIL PROTECTED] wrote: > Author: nextgens > Date: 2006-07-10 10:22:11 + (Mon, 10 Jul 2006) > New Revision: 9540 > > Log: > * More consistent locking in freenet/client/async, following lint4j > recommendations > * some log messages have switched from

Re: [freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
Hmmm, on second thoughts, lets not take this metaphor too far... if opennet mode is promiscuous, then darknet is polyamorous... Nonetheless it's a well established term in networking, so I think we should use it. On Tue, Jul 11, 2006 at 12:58:22PM +0100, Matthew Toseland wrote: > On Mon, Jul 10,

Re: [freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
On Tue, Jul 11, 2006 at 09:45:48AM +0200, Florent Daigni?re (NextGen$) wrote: > > > > Darknet is more secure than opennet. We should make this as clear as > > possible. Obviously we should make the warning as concise as possible; > > we could shrink the above down to "Warning: Opennet mode detecte

Re: [freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
On Mon, Jul 10, 2006 at 07:20:36PM -0700, Ian Clarke wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > Or simply call it "promiscuous mode", rather than "opennet mode" - I > think this captures the dangers quite concisely (the analogy to > promiscuous sexual behavior is a pretty goo

[freenet-dev] Opennet warning?

2006-07-11 Thread Matthew Toseland
othing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/a77ace33/attachment.pgp>

[freenet-dev] Re: [freenet-cvs] r9529 - trunk/freenet/src/freenet/node

2006-07-11 Thread Matthew Toseland
onWaiter() { > + Thread t; > synchronized(this) { > if(cw == null) > cw = new CompletionWaiter(); > else > return; > + t = new Thread(cw, "Completion waiter for "+uid); > } > - Thread t = new Thread(cw, "Completion waiter for "+uid); > t.setDaemon(true); > t.start(); > } > @@ -813,7 +821,7 @@ > } > } > > - public boolean completed() { > + public synchronized boolean completed() { > return allTransfersCompleted; > } -- Matthew J Toseland - toad at amphibian.dyndns.org Freenet Project Official Codemonkey - http://freenetproject.org/ ICTHUS - Nothing is impossible. Our Boss says so. -- next part -- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: <https://emu.freenetproject.org/pipermail/devl/attachments/20060711/9831b037/attachment.pgp>

Re: [freenet-dev] Opennet warning?

2006-07-11 Thread NextGen$
* Matthew Toseland <[EMAIL PROTECTED]> [2006-07-11 03:16:03]: > We could have a warning on the front page when running opennet, like > with testnet: > > "You are running freenet in opennet mode. While in opennet mode, you have > no control over who your node connects to. Note that those nodes you