Menu
▾
▴
sipdht-devel — This list is for development of sipdht project
You can subscribe to this list here.
| 2006 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(2) |
Jun
(10) |
Jul
(10) |
Aug
|
Sep
|
Oct
(2) |
Nov
(6) |
Dec
(1) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2007 |
Jan
(16) |
Feb
(14) |
Mar
|
Apr
(5) |
May
|
Jun
(4) |
Jul
(2) |
Aug
|
Sep
(3) |
Oct
|
Nov
(1) |
Dec
|
| 2008 |
Jan
(29) |
Feb
(22) |
Mar
|
Apr
|
May
|
Jun
|
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2013 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
(1) |
Dec
|
|
From: Enrico M. <enr...@te...> - 2006-11-27 15:13:49
|
Lorenzo wrote: > I'm a new entry in this mailing list, I discover your library today. I'm a Ph.D. student of University of Rome TorVergata. We're interested in developing a "Nat Friendly" SIP DHT, trying to mediate from the sipp2p > > mediate the draft-bryan-sipping-p2p-03 and the common Nat Traversal best pratices (draft-ietf-behave-p2p-state-00.txt). Do you already covered some of these NAT Traversal issues in your implementation? Actually we don't implement neither version 03 of the draft nor any NAT traversal mechanism. However, it's our intent to start some experimental branches on the repository for investigating other DHT protocols, other than the trivial one currently implemented. Needless to say, NAT traversal will be addressed as one of the main issues. The structure of the code should easily allow to interchange the DHT protocol; you can try it by yourself (and in case let us know), or just stay tuned and wait for some news in the near future :-) -- Ciao, Enrico |
|
From: Lorenzo <zid...@au...> - 2006-11-27 12:29:45
|
Hi, I'm a new entry in this mailing list, I discover your library today. I'm a Ph.D. student of University of Rome TorVergata. We're interested in developing a "Nat Friendly" SIP DHT, trying to mediate from the sipp2p mediate the draft-bryan-sipping-p2p-03 and the common Nat Traversal best pratices (draft-ietf-behave-p2p-state-00.txt). Do you already covered some of these NAT Traversal issues in your implementation? Good Work! Lorenzo |
|
From: Enrico M. <enr...@te...> - 2006-11-20 13:19:09
|
Folks, recently we've been working hard for letting application written in languages other than C use libsipdht. Eventually we have released bindings for C++, Java and Python, plus many examples (also a ~200-lines graphical application written in python) and some docs. Moreover, we have added support for letting peer, put and get objects share the same nta_agent (i.e. use the same socket); unfortunately, slight changes to the API were required, so the new version 0.5.x isn't backward compatible. If you are interested, give a look at the brand new modules libsipdht++, libsipdht-java and libsipdht-python. -- Ciao, Enrico |
|
From: Enrico M. <enr...@te...> - 2006-11-09 17:36:37
|
Ludovico Cavedon wrote: > I have two questions. > The first could be a bug. In file put.c, at the end of callback > put_res_handler(), line 122, there is: > su_home_unref(put->p_home); > So it deallocates the memory where the put object is stored, > which was allocated in sipdht_put_create() at line 70: > put = su_home_new(sizeof(sipdht_put_t)); > > So, when later someone calls sipdht_put_deinit() and sipdht_put_destroy(), > they act on the put object which is no longer valid. > Moreover sipdht_put_destroy() calls: > su_home_unref(put->p_home); > > What do you think about it? Maybe line 122 should not be there? I cannot look at the sources now, but as far as I remember it _is_ a bug. > The same considerations apply to get.c. Indeed. > Second question: I am trying to make a client that uses sipdht, which > should act as a peer and perform GETs and PUTs within the same process. > > The easiest way it so glue together the code from sipdht-tools, but using > just one root object. This is not so nice because it opens a socket for > the peer, one for each put or get, and the process connects to itself via UDP. Currently it is the only way for letting a peer and a client coexist in the same application. We are working for providing a mechanism for letting them share the same port. > To use just one socket I guess I should create just one agent and create > different legs. So I cannot use sipdht public API, but internal one. Am I right? Try to avoid that: it would break compatibility even with bugfix releases.. -- Ciao, Enrico |
|
From: Ludovico C. <lud...@cs...> - 2006-11-09 09:44:16
|
Hi,
I am trying to play with SIPDHT.
I have two questions.
The first could be a bug. In file put.c, at the end of callback
put_res_handler(), line 122, there is:
su_home_unref(put->p_home);
So it deallocates the memory where the put object is stored,
which was allocated in sipdht_put_create() at line 70:
put = su_home_new(sizeof(sipdht_put_t));
So, when later someone calls sipdht_put_deinit() and sipdht_put_destroy(),
they act on the put object which is no longer valid.
Moreover sipdht_put_destroy() calls:
su_home_unref(put->p_home);
What do you think about it? Maybe line 122 should not be there?
The same considerations apply to get.c.
Second question: I am trying to make a client that uses sipdht, which
should act as a peer and perform GETs and PUTs within the same process.
The easiest way it so glue together the code from sipdht-tools, but using
just one root object. This is not so nice because it opens a socket for
the peer, one for each put or get, and the process connects to itself via UDP.
To use just one socket I guess I should create just one agent and create
different legs. So I cannot use sipdht public API, but internal one. Am I right?
Thank you for you attention,
Reagrds,
Ludovico Cavedon
|
|
From: Enrico M. <enr...@te...> - 2006-11-02 16:43:00
|
Hi all,
we have just added routing functionalities (instead of redirection) to
both the library and the tools.
Routing can be toggled through:
+ the sipdht_peer_set_routing function in the library;
+ the '-R' option in the sipdht tool.
The new functionalities are available in libsipdht 0.4.0 and
sipdht-tools 0.2.0.
Once again, many thanks to Stefano who contributed most of the code.
--=20
Ciao,
Enrico
Ahem.. Sorry, the notice below is not my fault :-(
--------------------------------------------------------------------
CONFIDENTIALITY NOTICE
This message and its attachments are addressed solely to the persons =
above and may contain confidential information. If you have received the =
message in error, be informed that any use of the content hereof is =
prohibited. Please return it immediately to the sender and delete the =
message. Should you have any questions, please contact us by replying to =
web...@te....
Thank you
www.telecomitalia.it
--------------------------------------------------------------------
=20
|
|
From: Enrico M. <enr...@te...> - 2006-10-19 10:02:09
|
We have just added a chat and a conference room.
The chat room is available at si...@co... and can be
joined from any jabber client (Gaim, PSI, GTalk...); developers are
often on line (during working ours in Central Europe, i.e. in the
morning if you are in America or in the evening if you are in Asia ;-) )
and that's the right place for satisfing any kind of curiosity about the
project.
The voice conference room is sip:si...@si... and for joining
it a SIP phone is required (Kphone, Linphone, X-Lite, or Wengophone
which works also when only HTTP connectivity is available). The room is
always active, but often empty; events happening there will be announced
on this list.
--=20
Ciao,
Enrico
Ahem.. Sorry, the notice below is not my fault :-(
--------------------------------------------------------------------
CONFIDENTIALITY NOTICE
This message and its attachments are addressed solely to the persons =
above and may contain confidential information. If you have received the =
message in error, be informed that any use of the content hereof is =
prohibited. Please return it immediately to the sender and delete the =
message. Should you have any questions, please contact us by replying to =
web...@te....
Thank you
www.telecomitalia.it
--------------------------------------------------------------------
=20
|
|
From: Enrico M. <enr...@te...> - 2006-10-13 12:56:58
|
Folks: after much hard-coding and hard-writing we have completed the refactoring. Now the project has two modules: + libsipdht: a dynamic library with full documentation, meant to be used in external applications like softphones or SIP proxies; + sipdht-tools: the old collection of tools for playing with DHTs, reinmplemented using the library. The web page at sipdht.sf.net has been completely rewritten and Doxygen documentation (full of examples) added. Enjoy it if you wish! --=20 Ciao, Enrico Ahem.. Sorry, the notice below is not my fault :-( -------------------------------------------------------------------- CONFIDENTIALITY NOTICE This message and its attachments are addressed solely to the persons = above and may contain confidential information. If you have received the = message in error, be informed that any use of the content hereof is = prohibited. Please return it immediately to the sender and delete the = message. Should you have any questions, please contact us by replying to = web...@te.... Thank you www.telecomitalia.it -------------------------------------------------------------------- =20 |
|
From: Enrico M. <enr...@te...> - 2006-07-29 01:20:44
|
Version 0.2.0 is now available for downloading. It works fine with a clean version of sofia-sip libraries, but doesn't compile any more with any patched one. Now it is possible to have binary packages depending from "official" Sofia-SIP. -- Enrico |
|
From: Enrico M. <enr...@te...> - 2006-07-28 21:45:18
|
Kasthuri Ram V. wrote: > I have built an FC5 rpm for sipdht-0.1.1. Believe this would help many > who find it difficult to patch and build sofia-sip library as well as > sipdht. Also, availability of rpm packages would help people upgrade to > next version easily. > > Would like to know if sipdht's rpm and source rpm can be hosted at the > sourceforge site. Of course they can. We just have to understand how to deal with sofia-sip. I'd like to have one rpm for sipdht and one for a patched version of sofia-sip. The latter could then be hosted either on sofia-sip's project page or on sipdht's one. If you think you can build both rpms, we can get in touch with sofia-sip's developers. -- Enrico |
|
From: Enrico M. <enr...@te...> - 2006-07-28 04:53:01
|
Kasthuri Ram V wrote: > Milestone 1.3 is very important, because it allows to release sipdht > packages depending from official sofia-sip releases (also there, they > are spending some effort for generating binary distributions). I guess > I can apply for reaching 1.1 it in a couple of weeks (right now I'm very > busy in writing Internet drafts I can't delay :-( ). If we don't find > anyone willing to write man pages, I think we could skip 1.2 and release > version 0.2 with missing documentation. > > > If we can't get man pages, we may atleast have a readme available > through rpm or doc-rpm and place it in /usr/share/doc/sipdht* That seems the most affordable fallback solution. I hope to have some working code to avoid patching next week or something. -- Enrico |
|
From: Enrico M. <enr...@te...> - 2006-07-27 23:52:47
|
Kasthuri Ram V. wrote: > [Changed subject line to reflect relevance to present discussion]. > >> Great! If you agree to be responsible for rpm generation I can add you >> to the project's member list with such a role. Moreover, if you already >> have an rpm for a patched version of sofia-sip-1.12.0 (I'd name it >> sofia-sip-1.12.0-sipdht*.rpm), just put it online (or send it to me), >> and we'll show it to sofia-sip developers asking if they can put it on >> their page. > > Would be glad to be a member and would try my best to keep rpm > generation updated. Will be fine to have a package maintainer :-) I hope we'll also be able to release .deb packages, eventually. > To begin with, I have uploaded the rpms of patched version of > sofia-sip-1.12.0 and sipdht. The rpm and source rpm can be downloaded > from: > > http://www.freewebs.com/kasto/rpms/sofia-sip-sipdht-lib-0.1.1-1.i386.rpm > > http://www.freewebs.com/kasto/rpms/sofia-sip-sipdht-lib-0.1.1-1.src.rpm > [Note: While downloading the rpms on linux machine, browsers may try to > open the files as audio files because of the MIME type set by the > website. Appropriate settings may need to be used to download the files > or one may use 'wget' or 'lynx'] I hope these packages will be useful to many people. I've been thinking about the possible evolution of the project and I'd like to discuss what the best way forward would be. A mail with some comments will follow. -- Enrico |
|
From: Enrico M. <enr...@te...> - 2006-07-27 23:40:44
|
I've talked to Kai Vehmanen, a sofia-sip developer, and he suggested a
trick for icluding new headers without patching the stack. This has
made me think about a possible roadmap for the evolution of sipdht:
1.1. implementing the mechanism for run-time loading of DHT-* headers in
sofia-sip stack;
1.2. writing man pages for sipdht, sipdhtp, sipdhtg and sipdhtq (actual
documentation in README should be a good starting point);
1.3. releasing sipdht version 0.2, .tgz, .rpm and .src.rpm (and possibly
.deb);
2.1. splitting the project in two modules: libsipdht and sipdht-tools.
The former will be a dynamic loadable library usable in custom
applications, while the latter will include sipdht* tools;
2.2. writing doxygen documentation for library functions (actual
documentation in .h files should be a good starting point);
2.3. releasing libsipdht 0.1 and sipdht-tools 0.1, .tgz, .rpm,
.src.rpm and -doc.rpm (and possibly also .deb and -doc.deb).
Milestone 1.3 is very important, because it allows to release sipdht
packages depending from official sofia-sip releases (also there, they
are spending some effort for generating binary distributions). I guess
I can apply for reaching 1.1 it in a couple of weeks (right now I'm very
busy in writing Internet drafts I can't delay :-( ). If we don't find
anyone willing to write man pages, I think we could skip 1.2 and release
version 0.2 with missing documentation.
Then, I care much about milestone 2.3 because it would let easily
integrate sipdht in applications. I think we can get there in the first
half of September.
I would appreciate any comment and if you think you can help in any of
these items, you are welcome.
-- Enrico
|
|
From: Kasthuri R. V. <Kas...@in...> - 2006-07-27 20:17:29
|
[Changed subject line to reflect relevance to present discussion]. > Great! If you agree to be responsible for rpm generation I can add you > to the project's member list with such a role. Moreover, if you already > have an rpm for a patched version of sofia-sip-1.12.0 (I'd name it > sofia-sip-1.12.0-sipdht*.rpm), just put it online (or send it to me), > and we'll show it to sofia-sip developers asking if they can put it on > their page. Would be glad to be a member and would try my best to keep rpm generation updated. To begin with, I have uploaded the rpms of patched version of sofia-sip-1.12.0 and sipdht. The rpm and source rpm can be downloaded from: http://www.freewebs.com/kasto/rpms/sofia-sip-sipdht-lib-0.1.1-1.i386.rpm http://www.freewebs.com/kasto/rpms/sofia-sip-sipdht-lib-0.1.1-1.src.rpm [Note: While downloading the rpms on linux machine, browsers may try to open the files as audio files because of the MIME type set by the website. Appropriate settings may need to be used to download the files or one may use 'wget' or 'lynx'] sofia-sip-sipdht-lib rpm would provide the following files: /usr/lib/libsofia-sip-ua.a /usr/lib/libsofia-sip-ua.la /usr/lib/libsofia-sip-ua.so /usr/lib/libsofia-sip-ua.so.0 /usr/lib/libsofia-sip-ua.so.0.0.0 Details of the rpm is as below: [root@~] rpm -qai sofia-sip-sipdht-lib Name : sofia-sip-sipdht-lib Relocations: (not relocatable) Version : 0.1.1 Vendor: (none) Release : 1 Build Date: Wed 26 Jul 2006 06:44:52 PM IST Install Date: Thu 27 Jul 2006 12:00:16 PM IST Build Host: kasto-machine Group : Library Source RPM: sofia-sip-sipdht-lib-0.1 .1-1.src.rpm Size : 10033132 License: LGPL Signature : (none) Packager : Kasto <kas...@gm...> Summary : Sofia SIP library with a patch for sipdht Description : Sofia SIP library version 1.12.0 with a patch for sipdht version 0.1.1 Please do let me know if any modifications are needed. Thanks, Kasto -- **************** CAUTION - Disclaimer ***************** This e-mail contains PRIVILEGED AND CONFIDENTIAL INFORMATION intended solely for the use of the addressee(s). If you are not the intended recipient, please notify the sender by e-mail and delete the original message. Further, you are not to copy, disclose, or distribute this e-mail or its contents to any other person and any such actions are unlawful. This e-mail may contain viruses. Infosys has taken every reasonable precaution to minimize this risk, but is not liable for any damage you may sustain as a result of any virus in this e-mail. You should carry out your own virus checks before opening the e-mail or attachment. Infosys reserves the right to monitor and review the content of all messages sent to or from this e-mail address. Messages sent to or from this e-mail address may be stored on the Infosys e-mail system. ***INFOSYS******** End of Disclaimer ********INFOSYS*** |
|
From: Kasthuri R. V. <Kas...@in...> - 2006-07-27 17:43:57
|
Hi, I have built an FC5 rpm for sipdht-0.1.1. Believe this would help many who find it difficult to patch and build sofia-sip library as well as sipdht. Also, availability of rpm packages would help people upgrade to next version easily. Would like to know if sipdht's rpm and source rpm can be hosted at the sourceforge site. Thanks, Kasto -- On Tue, 2006-07-25 at 14:08 +0200, Enrico Marocco wrote: > Folks, > > sipdht version 0.1.1 has just been released. New version is eventually > tolerant to multiple faults: the overlay seems now pretty robust. Still > misses data replication. > > Enrico > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys -- and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Sipdht-devel mailing list > Sip...@li... > https://lists.sourceforge.net/lists/listinfo/sipdht-devel **************** CAUTION - Disclaimer ***************** This e-mail contains PRIVILEGED AND CONFIDENTIAL INFORMATION intended solely for the use of the addressee(s). If you are not the intended recipient, please notify the sender by e-mail and delete the original message. Further, you are not to copy, disclose, or distribute this e-mail or its contents to any other person and any such actions are unlawful. This e-mail may contain viruses. Infosys has taken every reasonable precaution to minimize this risk, but is not liable for any damage you may sustain as a result of any virus in this e-mail. You should carry out your own virus checks before opening the e-mail or attachment. Infosys reserves the right to monitor and review the content of all messages sent to or from this e-mail address. Messages sent to or from this e-mail address may be stored on the Infosys e-mail system. ***INFOSYS******** End of Disclaimer ********INFOSYS*** |
|
From: Kasthuri R. V <kas...@gm...> - 2006-07-27 12:09:16
|
> 2.2. writing doxygen documentation for library functions (actual > documentation in .h files should be a good starting point); Cool! 2.3. releasing libsipdht 0.1 and sipdht-tools 0.1, .tgz, .rpm, > .src.rpm and -doc.rpm (and possibly also .deb and -doc.deb). Haven't tried my hands at .deb packages. Just got introduced to debian through Ubuntu. Would give it a try soon. Milestone 1.3 is very important, because it allows to release sipdht > packages depending from official sofia-sip releases (also there, they > are spending some effort for generating binary distributions). I guess > I can apply for reaching 1.1 it in a couple of weeks (right now I'm very > busy in writing Internet drafts I can't delay :-( ). If we don't find > anyone willing to write man pages, I think we could skip 1.2 and release > version 0.2 with missing documentation. If we can't get man pages, we may atleast have a readme available through rpm or doc-rpm and place it in /usr/share/doc/sipdht* Thanks, Kasto -- |
|
From: Kasthuri R. V. <Kas...@in...> - 2006-07-26 09:18:02
|
Hi Enrico, > Of course they can. We just have to understand how to deal with > sofia-sip. I'd like to have one rpm for sipdht and one for a patched > version of sofia-sip. The latter could then be hosted either on > sofia-sip's project page or on sipdht's one. Was thinking on similar lines. > If you think you can build both rpms, we can get in touch with > sofia-sip's developers. Sure. Thanks, Kasto -- > > -- Enrico **************** CAUTION - Disclaimer ***************** This e-mail contains PRIVILEGED AND CONFIDENTIAL INFORMATION intended solely for the use of the addressee(s). If you are not the intended recipient, please notify the sender by e-mail and delete the original message. Further, you are not to copy, disclose, or distribute this e-mail or its contents to any other person and any such actions are unlawful. This e-mail may contain viruses. Infosys has taken every reasonable precaution to minimize this risk, but is not liable for any damage you may sustain as a result of any virus in this e-mail. You should carry out your own virus checks before opening the e-mail or attachment. Infosys reserves the right to monitor and review the content of all messages sent to or from this e-mail address. Messages sent to or from this e-mail address may be stored on the Infosys e-mail system. ***INFOSYS******** End of Disclaimer ********INFOSYS*** |
|
From: Enrico M. <enr...@te...> - 2006-07-25 12:07:46
|
Folks, sipdht version 0.1.1 has just been released. New version is eventually tolerant to multiple faults: the overlay seems now pretty robust. Still misses data replication. Enrico |
|
From: Mosiuoa T. <g06...@ca...> - 2006-06-30 08:59:01
|
Hi again, Sorry, the URL for those interested is http://open-chord.sourceforge.net Cheers! -----Original Message----- From: sip...@li... [mailto:sip...@li...] On Behalf Of sip...@li... Sent: 29 June 2006 21:42 To: sip...@li... Subject: Sipdht-devel Digest, Vol 1, Issue 6 Send Sipdht-devel mailing list submissions to sip...@li... To subscribe or unsubscribe via the World Wide Web, visit https://lists.sourceforge.net/lists/listinfo/sipdht-devel or, via email, send a message with subject or body 'help' to sip...@li... You can reach the person managing the list at sip...@li... When replying, please edit your Subject line so it is more specific than "Re: Contents of Sipdht-devel digest..." Today's Topics: 1. Chord Project (Mosiuoa Tsietsi) ---------------------------------------------------------------------- Message: 1 Date: Thu, 29 Jun 2006 17:50:18 +0200 From: "Mosiuoa Tsietsi" <g06...@ca...> Subject: [Sipdht-devel] Chord Project To: <sip...@li...> Message-ID: <000001c69b93$b88cf0a0$377...@ic...> Content-Type: text/plain; charset="us-ascii" Hi. I thought you might want to know of another Chord project which is being run on sourceforge called OpenChord that has released an open source implementation (in Java) of the Chord routing substrate. There is a lot of abstraction layered into the architecture and allows developers to layer applications over it, by providing an API to interact with the lower abstraction layers. I managed to write an app over OpenChord in just a few hours. It's worth a try. Cheers! -------------- next part -------------- An HTML attachment was scrubbed... URL: http://sourceforge.net/mailarchive/forum.php?forum=sipdht-devel/attachments/ 20060629/f29a0c71/attachment.html ------------------------------ Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 ------------------------------ _______________________________________________ Sipdht-devel mailing list Sip...@li... https://lists.sourceforge.net/lists/listinfo/sipdht-devel End of Sipdht-devel Digest, Vol 1, Issue 6 ****************************************** |
|
From: Mosiuoa T. <g06...@ca...> - 2006-06-29 15:51:53
|
Hi. I thought you might want to know of another Chord project which is being run on sourceforge called OpenChord that has released an open source implementation (in Java) of the Chord routing substrate. There is a lot of abstraction layered into the architecture and allows developers to layer applications over it, by providing an API to interact with the lower abstraction layers. I managed to write an app over OpenChord in just a few hours. It's worth a try. Cheers! |
|
From: Enrico M. <enr...@te...> - 2006-06-09 12:39:35
|
Mosiuoa Tsietsi wrote: > I have been testing out the sipdht-0.0.2 release. Its working fine, but I have a query about the finger table entries that are generated when I simlulate nodes joining the overlay (DHT). The DHT surely is based on CHORD since it uses finger tables (and draft-bryan-sipping-p2p-02 uses CHORD as example). Actually the DHT implemented so far is not properly chord, since the finger table management is simplified. In fact it is populated using only the information retrieved from the successor: basically, when the DHT is stabilized, finger `i', which, for i > 0, should be the responsible for the right-open interval [n + 2^(i-1), n + 2^1[, will be populated by the closer _known_ node to the start of the interval. This method, which actually isn't documented :-(, permits a drastical reduction in the number of messages exchanged during the stabilization procedure. On the other hand, the number of iterations for get/put seems to increase in the worst case from log2N to log4N. > Say three nodes (127.0.1.1, 127.0.1.2 and 127.0.1.3) join the overlay (in that order). And say [2] bootstraps through [1] and [3] through [2]. Using the sipdhtq command for displaying finger tables of the nodes with hash cardinality of 8, I get for [1]: > > Finger 0: (0x2) : sip:2@127.0.1.1 Right. > Finger 1: (0x3) : sip:2@127.0.1.1 The interval is [2, 3[ and the node closest to 2 is 1; since 1 is the actual node, it can be replaced by its successor. > Finger 2: (0x5) : sip:2@127.0.1.1 The interval is [3, 5[ and the node closest to 3 is 2. If the cardinality had been 16, the 4th finger, responsible of [5, 9[ would have been 3. If it is not clear, try to give a look to the sipdht_fit_finger function defined in dht_util.c (its description is in dht_util.h). > ....... > > Finger 0 gives the immediate successor of the node in the circular space, which I understand. [2] is the immediate successor for [1], ok. But I don't understand the i^th entries for the finger tables of [1]. Using the formula s = successor(n + 2^i-1 ) formula, I should get for [1]: > > Finger 0: (0x2) : sip:2@127.0.1.1 > Finger 1: (0x3) : sip:3@127.0.1.1 > Finger 2: (0x5) : sip:1@127.0.1.1 > ....... > > Please help me out if I got my understanding of CHORD mixed up, or I'm failing to interpret the output correctly. None of the above. > Again, if I use cardinality of 8, I also thought it would only give three lines of output (m=3 for three bit identifiers if there are 8 nodes: 0-7) for each node's finger tables. It is. Do not confuse hash cardinality (defined as SIPDHT_MAX_N in dht.h) with sipdhtq -N parameter which is used only for output formatting. Enrico |
|
From: Mosiuoa T. <g06...@ca...> - 2006-06-09 11:35:15
|
Hi all! I have been testing out the sipdht-0.0.2 release. Its working fine, but I have a query about the finger table entries that are generated when I simlulate nodes joining the overlay (DHT). The DHT surely is based on CHORD since it uses finger tables (and draft-bryan-sipping-p2p-02 uses CHORD as example). Say three nodes (127.0.1.1, 127.0.1.2 and 127.0.1.3) join the overlay (in that order). And say [2] bootstraps through [1] and [3] through [2]. Using the sipdhtq command for displaying finger tables of the nodes with hash cardinality of 8, I get for [1]: Finger 0: (0x2) : sip:2@127.0.1.1 Finger 1: (0x3) : sip:2@127.0.1.1 Finger 2: (0x5) : sip:2@127.0.1.1 ....... Finger 0 gives the immediate successor of the node in the circular space, which I understand. [2] is the immediate successor for [1], ok. But I don't understand the i^th entries for the finger tables of [1]. Using the formula s = successor(n + 2^i-1 ) formula, I should get for [1]: Finger 0: (0x2) : sip:2@127.0.1.1 Finger 1: (0x3) : sip:3@127.0.1.1 Finger 2: (0x5) : sip:1@127.0.1.1 ....... Please help me out if I got my understanding of CHORD mixed up, or I'm failing to interpret the output correctly. Again, if I use cardinality of 8, I also thought it would only give three lines of output (m=3 for three bit identifiers if there are 8 nodes: 0-7) for each node's finger tables. Thanks again! |
|
From: Mosiuoa T. <g06...@ca...> - 2006-06-07 15:05:00
|
Eurika! Got it working. It was the linking alright. Thanks! I plan to do alot of testing from now on. Forgot to mention that I'm a student reading for a MSc in South Africa. My research is on P2P SIP and I'm hoping to make a significant contribution in the area, particularly when it comes to cross overlay communications involving two overlays using different DHT algorithms ie (Chord-Pastry). Cheers. Enrico Marocco wrote: >Mosiuoa Tsietsi wrote: > > >>When I attempt to run sipdht using the : >>sipdht -a address -s interval >>syntax, it says: >> >>[begin] >>./src/sipdht: error while loading shared libraries: >>libsofia-sip-ua.so.0: cannot open shared object file: No such file or >>directory >>[end] >> >> > >You need to teel the linker where sofia-sip libraries are installed. >This can be done setting the LD_LIBRARY_PATH variable. For example, if >you have installed sofia-sip in /usr/local/lib, before running sipdht, do: > >$ export LD_LIBRARY_PATH=/usr/local/lib > >Enrico > > > |
|
From: Enrico M. <enr...@te...> - 2006-06-07 10:33:39
|
Mosiuoa Tsietsi wrote: > When I attempt to run sipdht using the : > sipdht -a address -s interval > syntax, it says: > > [begin] > ./src/sipdht: error while loading shared libraries: > libsofia-sip-ua.so.0: cannot open shared object file: No such file or > directory > [end] You need to teel the linker where sofia-sip libraries are installed. This can be done setting the LD_LIBRARY_PATH variable. For example, if you have installed sofia-sip in /usr/local/lib, before running sipdht, do: $ export LD_LIBRARY_PATH=/usr/local/lib Enrico |
|
From: Mosiuoa T. <g06...@ca...> - 2006-06-07 10:29:35
|
Hi all! Thanks for the help. I used the export command as instrcuted, and sipdht configured well and I ran a make on it as well which succeeded. When I attempt to run sipdht using the : sipdht -a address -s interval syntax, it says: [begin] ./src/sipdht: error while loading shared libraries: libsofia-sip-ua.so.0: cannot open shared object file: No such file or directory [end] I am using the sipdht-0.0.2 release with sofia-sip-1.11.8 stack. I applied the patches to sofia-sip-1.11.8 and ran *./configure* and *make* and *make install* commands successfully. Both sipdht and sofia-sip are in the /usr/local/src folder. As always. any help will be appreciated! Cheers! Enrico Marocco wrote: >Mosiuoa Tsietsi wrote: > > >>Hi. I downloaded the sipdht software a while ago and have been trying >>to get it running on my machine. When I execute the ./configure >>command, it proceeds well but ends up saying the following: >> >>"checking for SOFIA_SIP_UA... configure: error: Package requirements >>(sofia-sip-ua) were not met: >> >>No package 'sofia-sip-ua' found >> >>Consider adjusting the PKG_CONFIG_PATH environment variable if you >>installed software in a non-standard prefix. >> >> > >Have you ran `make install' in sofia-sip directory? If yes, it means >that you have to explicitly tell pkg-config where to find sofia-sip pkg >description file. Say you have installed sofia-sip libraries with >`prefix' set to `/usr/local', try: > >$ export PKG_CONFIG_PATH=/usr/local/lib/pkgconfig >$ cd <path-to-sipdht> >$ ./configure > >Enrico > > > |
