From yunlzy at 126.com Sun Mar 1 07:50:50 2015 From: yunlzy at 126.com (=?GBK?B?093B+uzP?=) Date: Sun, 1 Mar 2015 23:50:50 +0800 (CST) Subject: [ndnSIM] simulate an IP-overlay network Message-ID: <4f6ff018.47b.14bd6086a65.Coremail.yunlzy@126.com> Hi: I want to simulate an scenario with NDN as an ip-overlay network, which need to send package between ip and ndn nodes(that's the only way i can find to simulate an ip-overlay network). Can ndnSim do this? I know in ndnSIM1.0 a class named IpFacesHelper can help me with this problem, but what can i do in ndnSIM2.0? Or is there any better idea for me to do a simulation with ip-overlay scenario? Thanks very much. Regards, Cobb -------------- next part -------------- An HTML attachment was scrubbed... URL: From anselme333 at gmail.com Sun Mar 1 08:45:19 2015 From: anselme333 at gmail.com (NDIKUMANA ANSELME) Date: Mon, 2 Mar 2015 01:45:19 +0900 Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node Message-ID: Hi, For preventing the buffer overflow, I would like to setup the notification message, in which the intermediate node generates notification message (on buffer threshold) to its downstream node. Any idea on how Intermediate node can generate notification message to its downstream node, will be appreciated. Thank you in advance for your help. Anselme N. -------------- next part -------------- An HTML attachment was scrubbed... URL: From spiros.mastorakis at gmail.com Sun Mar 1 13:12:34 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Sun, 1 Mar 2015 13:12:34 -0800 Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node In-Reply-To: References: Message-ID: <0B341044-648E-469B-A0C2-03088707C4BF@gmail.com> Hello, interesting question... I think that a convenient way to do so is to define an application that will run on the intermediate nodes and will send notifications to a downstream node based on the desired conditions. At least, this would be the way that I would start implementing it. Another way to do so (I do not think that it is quite convenient though) is to change the forwarding pipelines of NFD so that to send notifications to a downstream node based on the desired conditions. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 1, 2015, at 8:45 AM, NDIKUMANA ANSELME wrote: > > Hi, > > For preventing the buffer overflow, I would like to setup the notification message, in which the intermediate node generates notification message (on buffer threshold) to its downstream node. > > Any idea on how Intermediate node can generate notification message to its downstream node, will be appreciated. > > > > Thank you in advance for your help. > > > > Anselme N. > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From anselme333 at gmail.com Sun Mar 1 16:26:08 2015 From: anselme333 at gmail.com (NDIKUMANA ANSELME) Date: Mon, 2 Mar 2015 09:26:08 +0900 Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node In-Reply-To: <0B341044-648E-469B-A0C2-03088707C4BF@gmail.com> References: <0B341044-648E-469B-A0C2-03088707C4BF@gmail.com> Message-ID: Dear Sir, Thanks for the feedback. i defined the application that run on intermediate nodes but the only problem i have , it forwards the notification to upstream node instead of downstream node. With Regards, Anselme N. 2015-03-02 6:12 GMT+09:00 Spyridon (Spyros) Mastorakis < spiros.mastorakis at gmail.com>: > Hello, > > interesting question... > > I think that a convenient way to do so is to define an application that > will run on the intermediate nodes and will send notifications to a > downstream node based on the desired conditions. At least, this would be > the way that I would start implementing it. > > Another way to do so (I do not think that it is quite convenient though) > is to change the forwarding pipelines of NFD so that to send notifications > to a downstream node based on the desired conditions. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > > > On Mar 1, 2015, at 8:45 AM, NDIKUMANA ANSELME > wrote: > > > > Hi, > > > > For preventing the buffer overflow, I would like to setup the > notification message, in which the intermediate node generates notification > message (on buffer threshold) to its downstream node. > > > > Any idea on how Intermediate node can generate notification message to > its downstream node, will be appreciated. > > > > > > > > Thank you in advance for your help. > > > > > > > > Anselme N. > > > > _______________________________________________ > > ndnSIM mailing list > > ndnSIM at lists.cs.ucla.edu > > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From amin1431 at yahoo.com Sun Mar 1 16:44:45 2015 From: amin1431 at yahoo.com (amin1431) Date: Mon, 2 Mar 2015 00:44:45 +0000 (UTC) Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node In-Reply-To: <54F3B189.7090807@ac.upc.edu> References: <54F3B189.7090807@ac.upc.edu> Message-ID: <1856468102.1055481.1425257085398.JavaMail.yahoo@mail.yahoo.com> Hi Anselme,Could you please share your code and scenario executed at ndnSIM for educational purposes? Thank you in advance. Sincerely, Armin | Subject: | Re: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node | | Date: | Mon, 2 Mar 2015 09:26:08 +0900 | | From: | NDIKUMANA ANSELME | | To: | Spyridon (Spyros) Mastorakis | | CC: | ndnsim | Dear Sir, Thanks for the feedback. i defined the application that run on intermediate nodes but the only problem i have , it forwards the notification to upstream node instead of downstream node. With Regards, Anselme N. 2015-03-02 6:12 GMT+09:00 Spyridon (Spyros) Mastorakis : Hello, interesting question... I think that a convenient way to do so is to define an application that will run on the intermediate nodes and will send notifications to a downstream node based on the desired conditions. At least, this would be the way that I would start implementing it. Another way to do so (I do not think that it is quite convenient though) is to change the forwarding pipelines of NFD so that to send notifications to a downstream node based on the desired conditions. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 1, 2015, at 8:45 AM, NDIKUMANA ANSELME wrote: > > Hi, > > For preventing the buffer overflow, I would like to setup the notification message, in which the intermediate node generates notification message (on buffer threshold) to its downstream node. > > Any idea on how Intermediate node can generate notification message to its downstream node, will be appreciated. > > > > Thank you in advance for your help. > > > > Anselme N. > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From amin1431 at yahoo.com Sun Mar 1 16:51:19 2015 From: amin1431 at yahoo.com (amin1431) Date: Mon, 2 Mar 2015 00:51:19 +0000 (UTC) Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <54F3B2B3.6090809@ac.upc.edu> References: <54F3B2B3.6090809@ac.upc.edu> Message-ID: <746337952.1097064.1425257479501.JavaMail.yahoo@mail.yahoo.com> Hi Spyros,Thank you for your help.Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? Sincerely, Armin ----------------------------------------------------------------------------------------------------------- | Subject: | Re: [ndnSIM] How make a inter-domain network by ndnSIM? | | Date: | Wed, 4 Feb 2015 18:06:38 -0800 | | From: | Spyridon (Spyros) Mastorakis | | To: | amin1431 | | CC: | Ndnsim | Hello, NDN can run on top of every protocol and ndnSIM implements this primitive. Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Feb 3, 2015, at 6:21 AM, amin1431 wrote: > > Hi friends :-) > I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. > > Could you please give me some hints, how does it possible? > > > Sincerely, > Armin > _______________________________________________ ndnSIM mailing list ndnSIM at lists.cs.ucla.edu http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From anilj.mailing at gmail.com Sun Mar 1 18:43:50 2015 From: anilj.mailing at gmail.com (Anil Jangam) Date: Sun, 1 Mar 2015 18:43:50 -0800 Subject: [ndnSIM] Assert in TypeId::LookupByName: error. Could not find the class. Message-ID: Hi, When we are using AppHelper to install applications on a ndn Node, which application classes are supported and how? For example, in one example code, following few lines installs a Consumer application on the ndn node. 86 ndn::AppHelper consumerHelper ("ns3::ndn::ConsumerCbr"); 87 consumerHelper.SetPrefix (prefix); 89 consumerHelper.Install (consumerNodes); However, I do not see how the ndn-consumer-cbr.h/.cc files are getting compiled? I am trying to add a new type of application by defining its own class. I think I have coded it right (e.g. TypeId method, and calling NS_LOG_COMPONENT_DEFINE ("ndn.SampleApp"), NS_OBJECT_ENSURE_REGISTERED (SampleApp); macros. 28 TypeId 29 SampleApp::GetTypeId (void) 30 { 31 static TypeId tid = TypeId ("ns3::ndn::SampleApp") 32 .SetGroupName ("Ndn") 33 .SetParent () 34 .AddConstructor () 35 ; 36 return tid; 37 } But I consistently get the below error. assert failed. cond="uid != 0", msg="Assert in TypeId::LookupByName: ns3::ndn::SampleApp not found", file=../src/core/model/type-id.cc, line=539 terminate called without an active exception Do we have include this file into (which?) any of the wscript to be able to compile it? /anil. -------------- next part -------------- An HTML attachment was scrubbed... URL: From anselme333 at gmail.com Mon Mar 2 02:58:08 2015 From: anselme333 at gmail.com (NDIKUMANA ANSELME) Date: Mon, 2 Mar 2015 19:58:08 +0900 Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node In-Reply-To: <1856468102.1055481.1425257085398.JavaMail.yahoo@mail.yahoo.com> References: <54F3B189.7090807@ac.upc.edu> <1856468102.1055481.1425257085398.JavaMail.yahoo@mail.yahoo.com> Message-ID: Dear Armin, I have to fix some errors inside my code and i hope by end of this week i can share them. With Regards, Anselme NDIKUMANA 2015-03-02 9:44 GMT+09:00 amin1431 : > Hi Anselme, > Could you please share your code and scenario executed at ndnSIM for > educational purposes? > > Thank you in advance. > > Sincerely, > Armin > > > > Subject: Re: [ndnSIM] ndnSIM 2.O Intermediate node generates > notification message to its downstream node Date: Mon, 2 Mar 2015 > 09:26:08 +0900 From: NDIKUMANA ANSELME > To: Spyridon (Spyros) Mastorakis > CC: ndnsim > > > Dear Sir, > > Thanks for the feedback. i defined the application that run on > intermediate nodes but the only problem i have , it forwards the > notification to upstream node instead of downstream node. > With Regards, > Anselme N. > > 2015-03-02 6:12 GMT+09:00 Spyridon (Spyros) Mastorakis < > spiros.mastorakis at gmail.com>: > > Hello, > > interesting question... > > I think that a convenient way to do so is to define an application that > will run on the intermediate nodes and will send notifications to a > downstream node based on the desired conditions. At least, this would be > the way that I would start implementing it. > > Another way to do so (I do not think that it is quite convenient though) > is to change the forwarding pipelines of NFD so that to send notifications > to a downstream node based on the desired conditions. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > > > On Mar 1, 2015, at 8:45 AM, NDIKUMANA ANSELME > wrote: > > > > Hi, > > > > For preventing the buffer overflow, I would like to setup the > notification message, in which the intermediate node generates notification > message (on buffer threshold) to its downstream node. > > > > Any idea on how Intermediate node can generate notification message to > its downstream node, will be appreciated. > > > > > > > > Thank you in advance for your help. > > > > > > > > Anselme N. > > > > _______________________________________________ > > ndnSIM mailing list > > ndnSIM at lists.cs.ucla.edu > > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From contez at gmail.com Mon Mar 2 08:36:40 2015 From: contez at gmail.com (Riccardo Loti) Date: Mon, 02 Mar 2015 17:36:40 +0100 Subject: [ndnSIM] Question about disabled links coloring faces red Message-ID: <54F49198.2070001@gmail.com> First of all hi everyone, being the first message on the mailing list I suppose it's the least courtesy for me to say it. ;) Second, and probably more important, I do have a question about the behaviour of ndnSIM on green-yellow-red nodes. I do have a custom application (which is basically a request n sequential packets so not important here I think) installed on a few nodes, a few others are producers and they are all randomly placed on a grid layout with PointToPoint connections, where all nodes have the basic ndn stack installer through the helper. We are still using (due to it being a long project) ndnSIM 1.0, Forwarding strategy is BestRoute, we use the GlobalRoutingHelper and invoke its CalculateAllPossibleRoutes before running the simulation, so not only the optimal paths are mapped, but all, as confirmed in the visualizer in the node's FIB. Now, we also implemented a Failer class which each n seconds fails all node links for m seconds (using the GetFace(x)->SetUp(false/true) methods for each face). The problem is that when the failed node is on the path between the consumer and the producer all interests stops at the failed link, without going to an alternative (originally suboptimal) path. Shouldn't the faces going to the interrupted path be marked red instead of yellow so the routing can choose the second best one? Doesn't ndnSIM manages this in some way more or less automatically? Does it changes bewtween ndndSIM 1.0 and 2.0? Am I completely off the road and have totally misunderstood how this works, or missed a very obvious part of the documentation? (If so I'm sorry for wasting your time, but please explain me how so) Thank you very much and I'll be waiting for pointers (eh eh, C joke ^^). Riccardo From spiros.mastorakis at gmail.com Mon Mar 2 08:44:58 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Mon, 2 Mar 2015 08:44:58 -0800 Subject: [ndnSIM] ndnSIM 2.O Intermediate node generates notification message to its downstream node In-Reply-To: References: <0B341044-648E-469B-A0C2-03088707C4BF@gmail.com> Message-ID: Hello, possible reason for that: Is the notification expressed as an interest packet? If so, you have to think about its name prefix. If there is a FIB entry to an upstream node for this name prefix, it is normal to be forwarded upstream. In this case, probably you need either to change name prefix and add manually a FIB entry for that name prefix, or manually install a longer prefix match FIB entry for the name prefix of your interest. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 1, 2015, at 4:26 PM, NDIKUMANA ANSELME wrote: > > Dear Sir, > > Thanks for the feedback. i defined the application that run on intermediate nodes but the only problem i have , it forwards the notification to upstream node instead of downstream node. > With Regards, > Anselme N. > > 2015-03-02 6:12 GMT+09:00 Spyridon (Spyros) Mastorakis : > Hello, > > interesting question... > > I think that a convenient way to do so is to define an application that will run on the intermediate nodes and will send notifications to a downstream node based on the desired conditions. At least, this would be the way that I would start implementing it. > > Another way to do so (I do not think that it is quite convenient though) is to change the forwarding pipelines of NFD so that to send notifications to a downstream node based on the desired conditions. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > > > On Mar 1, 2015, at 8:45 AM, NDIKUMANA ANSELME wrote: > > > > Hi, > > > > For preventing the buffer overflow, I would like to setup the notification message, in which the intermediate node generates notification message (on buffer threshold) to its downstream node. > > > > Any idea on how Intermediate node can generate notification message to its downstream node, will be appreciated. > > > > > > > > Thank you in advance for your help. > > > > > > > > Anselme N. > > > > _______________________________________________ > > ndnSIM mailing list > > ndnSIM at lists.cs.ucla.edu > > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim > > From spiros.mastorakis at gmail.com Mon Mar 2 08:47:02 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Mon, 2 Mar 2015 08:47:02 -0800 Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <746337952.1097064.1425257479501.JavaMail.yahoo@mail.yahoo.com> References: <54F3B2B3.6090809@ac.upc.edu> <746337952.1097064.1425257479501.JavaMail.yahoo@mail.yahoo.com> Message-ID: <17DE51E4-57C3-4AE8-A6B7-11DDF51D778B@gmail.com> Hello Armin, to the best of my knowledge, we have not published such a tutorial. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 1, 2015, at 4:51 PM, amin1431 wrote: > > Hi Spyros, > Thank you for your help. > Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? > > > > Sincerely, > Armin > > ----------------------------------------------------------------------------------------------------------- > Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? > Date: Wed, 4 Feb 2015 18:06:38 -0800 > From: Spyridon (Spyros) Mastorakis > To: amin1431 > CC: Ndnsim > > Hello, > > NDN can run on top of every protocol and ndnSIM implements this primitive. > Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: > http://cs.ucla.edu/~mastorakis/ > > Internet Research Laboratory > PhD Computer Science > UCLA > > > > > > On Feb 3, 2015, at 6:21 AM, amin1431 > > wrote: > > > > Hi friends :-) > > I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. > > > > Could you please give me some hints, how does it possible? > > > > > > Sincerely, > > Armin > > > > > _______________________________________________ > ndnSIM mailing list > > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim > > > > From lanwang at memphis.edu Mon Mar 2 08:56:57 2015 From: lanwang at memphis.edu (Lan Wang (lanwang)) Date: Mon, 2 Mar 2015 16:56:57 +0000 Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <17DE51E4-57C3-4AE8-A6B7-11DDF51D778B@gmail.com> References: <54F3B2B3.6090809@ac.upc.edu> <746337952.1097064.1425257479501.JavaMail.yahoo@mail.yahoo.com> <17DE51E4-57C3-4AE8-A6B7-11DDF51D778B@gmail.com> Message-ID: <4A4B37FA-0C6B-48C9-B950-A4ABAB670863@memphis.edu> Not sure exactly what Armin is looking for. The following tutorial is not for ndnSIM, but how to connect the nfd instances running on different machines (not directly connected): http://yoursunny.com/t/2014/nfd-connect/ Lan On Mar 2, 2015, at 10:47 AM, "Spyridon (Spyros) Mastorakis" wrote: > Hello Armin, > > to the best of my knowledge, we have not published such a tutorial. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > >> On Mar 1, 2015, at 4:51 PM, amin1431 wrote: >> >> Hi Spyros, >> Thank you for your help. >> Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? >> >> >> >> Sincerely, >> Armin >> >> ----------------------------------------------------------------------------------------------------------- >> Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? >> Date: Wed, 4 Feb 2015 18:06:38 -0800 >> From: Spyridon (Spyros) Mastorakis >> To: amin1431 >> CC: Ndnsim >> >> Hello, >> >> NDN can run on top of every protocol and ndnSIM implements this primitive. >> Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: >> http://cs.ucla.edu/~mastorakis/ >> >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >>> On Feb 3, 2015, at 6:21 AM, amin1431 >> >> wrote: >>> >>> Hi friends :-) >>> I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. >>> >>> Could you please give me some hints, how does it possible? >>> >>> >>> Sincerely, >>> Armin >>> >> >> >> _______________________________________________ >> ndnSIM mailing list >> >> ndnSIM at lists.cs.ucla.edu >> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim >> >> >> >> > > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From amin1431 at yahoo.com Mon Mar 2 09:45:45 2015 From: amin1431 at yahoo.com (amin1431) Date: Mon, 2 Mar 2015 17:45:45 +0000 (UTC) Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <4A4B37FA-0C6B-48C9-B950-A4ABAB670863@memphis.edu> References: <4A4B37FA-0C6B-48C9-B950-A4ABAB670863@memphis.edu> Message-ID: <1732779336.1752731.1425318345207.JavaMail.yahoo@mail.yahoo.com> Hi Lan,Thank you for your guidance.I want actually connect two clusters (such as two smart grids or two vehicle networks) through NDN. I am trying to get the fundamental understanding to how it is possible to connect two clusters via NDN. for instance, how change the traffic from clusters to the Interest/Data and in reverse. If there is any useful material, I am very thankful to share with me. Sincerely, Armin From: Lan Wang (lanwang) To: Spyridon (Spyros) Mastorakis Cc: amin1431 ; Ndnsim Sent: Monday, March 2, 2015 5:56 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Not sure exactly what Armin is looking for.? The following tutorial is not for ndnSIM, but how to connect the nfd instances running on different machines (not directly connected): http://yoursunny.com/t/2014/nfd-connect/ Lan On Mar 2, 2015, at 10:47 AM, "Spyridon (Spyros) Mastorakis" wrote: > Hello Armin, > > to the best of my knowledge, we have not published such a tutorial. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > >> On Mar 1, 2015, at 4:51 PM, amin1431 wrote: >> >> Hi Spyros, >> Thank you for your help. >> Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? >> >> >> >> Sincerely, >> Armin >> >> ----------------------------------------------------------------------------------------------------------- >> Subject:??? Re: [ndnSIM] How make a inter-domain network by ndnSIM? >> Date:??? Wed, 4 Feb 2015 18:06:38 -0800 >> From:??? Spyridon (Spyros) Mastorakis >> To:??? amin1431 >> CC:??? Ndnsim >> >> Hello, >> >> NDN can run on top of every protocol and ndnSIM implements this primitive. >> Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: >> http://cs.ucla.edu/~mastorakis/ >> >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >>> On Feb 3, 2015, at 6:21 AM, amin1431 >> >> wrote: >>> >>> Hi friends :-) >>> I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. >>> >>> Could you please give me some hints, how does it possible? >>> >>> >>> Sincerely, >>> Armin >>> >> >> >> _______________________________________________ >> ndnSIM mailing list >> >> ndnSIM at lists.cs.ucla.edu >> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim >> >> >> >> > > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From lanwang at memphis.edu Mon Mar 2 14:47:51 2015 From: lanwang at memphis.edu (Lan Wang (lanwang)) Date: Mon, 2 Mar 2015 22:47:51 +0000 Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <1732779336.1752731.1425318345207.JavaMail.yahoo@mail.yahoo.com> References: <4A4B37FA-0C6B-48C9-B950-A4ABAB670863@memphis.edu> <1732779336.1752731.1425318345207.JavaMail.yahoo@mail.yahoo.com> Message-ID: <35019046-0F0E-45D6-8504-57C2828E0F57@memphis.edu> Armin, Do you mean that the traffic in the clusters is not NDN traffic but you want to convert the traffic to NDN at the border routers? I'm not sure how to do this. For some applications (e.g., http like traffic), this may be possible. But it's application specific, not something you can do regardless of the application. Lan On Mar 2, 2015, at 11:45 AM, amin1431 > wrote: Hi Lan, Thank you for your guidance. I want actually connect two clusters (such as two smart grids or two vehicle networks) through NDN. I am trying to get the fundamental understanding to how it is possible to connect two clusters via NDN. for instance, how change the traffic from clusters to the Interest/Data and in reverse. If there is any useful material, I am very thankful to share with me. Sincerely, Armin ________________________________ From: Lan Wang (lanwang) > To: Spyridon (Spyros) Mastorakis > Cc: amin1431 >; Ndnsim > Sent: Monday, March 2, 2015 5:56 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Not sure exactly what Armin is looking for. The following tutorial is not for ndnSIM, but how to connect the nfd instances running on different machines (not directly connected): http://yoursunny.com/t/2014/nfd-connect/ Lan On Mar 2, 2015, at 10:47 AM, "Spyridon (Spyros) Mastorakis" > wrote: > Hello Armin, > > to the best of my knowledge, we have not published such a tutorial. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > >> On Mar 1, 2015, at 4:51 PM, amin1431 > wrote: >> >> Hi Spyros, >> Thank you for your help. >> Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? >> >> >> >> Sincerely, >> Armin >> >> ----------------------------------------------------------------------------------------------------------- >> Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? >> Date: Wed, 4 Feb 2015 18:06:38 -0800 >> From: Spyridon (Spyros) Mastorakis > >> To: amin1431 > >> CC: Ndnsim > >> >> Hello, >> >> NDN can run on top of every protocol and ndnSIM implements this primitive. >> Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: >> http://cs.ucla.edu/~mastorakis/ >> >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >>> On Feb 3, 2015, at 6:21 AM, amin1431 >> > >> wrote: >>> >>> Hi friends :-) >>> I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. >>> >>> Could you please give me some hints, how does it possible? >>> >>> >>> Sincerely, >>> Armin >>> >> >> >> _______________________________________________ >> ndnSIM mailing list >> >> ndnSIM at lists.cs.ucla.edu >> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim >> >> >> >> > > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From amin1431 at yahoo.com Mon Mar 2 15:07:01 2015 From: amin1431 at yahoo.com (amin1431) Date: Mon, 2 Mar 2015 23:07:01 +0000 (UTC) Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <35019046-0F0E-45D6-8504-57C2828E0F57@memphis.edu> References: <35019046-0F0E-45D6-8504-57C2828E0F57@memphis.edu> Message-ID: <1378392944.2125914.1425337621211.JavaMail.yahoo@mail.yahoo.com> Hi,Yes, I want to connect two different traffics through the NDN.You said for some applications such as http, it is possible. Is there guideline? Sincerely, Armin From: Lan Wang (lanwang) To: amin1431 Cc: Spyridon (Spyros) Mastorakis ; Ndnsim Sent: Monday, March 2, 2015 11:47 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Armin, Do you mean that the traffic in the clusters is not NDN traffic but you want to convert the traffic to NDN at the border routers? ?I'm not sure how to do this. ?For some applications (e.g., http like traffic), this may be possible. ?But it's application specific, not something you can do regardless of the application. ?? Lan On Mar 2, 2015, at 11:45 AM, amin1431 wrote: Hi Lan,Thank you for your guidance.I want actually connect two clusters (such as two smart grids or two vehicle networks) through NDN. I am trying to get the fundamental understanding to how it is possible to connect two clusters via NDN. for instance, how change the traffic from clusters to the Interest/Data and in reverse. If there is any useful material, I am very thankful to share with me. Sincerely, Armin From: Lan Wang (lanwang) To: Spyridon (Spyros) Mastorakis Cc: amin1431 ; Ndnsim Sent: Monday, March 2, 2015 5:56 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Not sure exactly what Armin is looking for.? The following tutorial is not for ndnSIM, but how to connect the nfd instances running on different machines (not directly connected): http://yoursunny.com/t/2014/nfd-connect/ Lan On Mar 2, 2015, at 10:47 AM, "Spyridon (Spyros) Mastorakis" wrote: > Hello Armin, > > to the best of my knowledge, we have not published such a tutorial. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > >> On Mar 1, 2015, at 4:51 PM, amin1431 wrote: >> >> Hi Spyros, >> Thank you for your help. >> Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? >> >> >> >> Sincerely, >> Armin >> >> ----------------------------------------------------------------------------------------------------------- >> Subject:??? Re: [ndnSIM] How make a inter-domain network by ndnSIM? >> Date:??? Wed, 4 Feb 2015 18:06:38 -0800 >> From:??? Spyridon (Spyros) Mastorakis >> To:??? amin1431 >> CC:??? Ndnsim >> >> Hello, >> >> NDN can run on top of every protocol and ndnSIM implements this primitive. >> Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: >> http://cs.ucla.edu/~mastorakis/ >> >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >>> On Feb 3, 2015, at 6:21 AM, amin1431 >> >> wrote: >>> >>> Hi friends :-) >>> I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. >>> >>> Could you please give me some hints, how does it possible? >>> >>> >>> Sincerely, >>> Armin >>> >> >> >> _______________________________________________ >> ndnSIM mailing list >> >> ndnSIM at lists.cs.ucla.edu >> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim >> >> >> >> > > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Mon Mar 2 17:45:12 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Tue, 3 Mar 2015 10:45:12 +0900 Subject: [ndnSIM] Question about producer. Message-ID: Hello. I ran ndn-congestion-alt-topo-plugin example with FibHelper. but it seems the producer doesn't send a data packet. 18 p1 257 appFace:// InData 5.24288e-12 5.4272e-12 0 0 18 p1 257 appFace:// OutData 0 0 0 0 Does anyone know why? Do I have to designate Fib routing path? ndn::FibHelper::AddRoute("c1", "/data", "n1", 1); // link to n1 ndn::FibHelper::AddRoute("c2", "/data", "n1", 1); // link to n1 ndn::FibHelper::AddRoute("c3", "/data", "n1", 1); // link to n1 ndn::FibHelper::AddRoute("c4", "/data", "n1", 1); // link to n1 //ndn::FibHelper::AddRoute("n1", "/data", "n2", 1); // link to n2 ndn::FibHelper::AddRoute("n1", "/data", "n12", 2); // link to n12 ndn::FibHelper::AddRoute("n12", "/data", "n2", 1); // link to n2 ndn::FibHelper::AddRoute("n2", "/data/p1", "p1", 1); // link to p1 //ndn::FibHelper::AddRoute("n2", "/data/p2", "p2", 1); // link to p2 //ndn::FibHelper::AddRoute("n2", "/data/p3", "p3", 1); // link to p3 //ndn::FibHelper::AddRoute("n2", "/data/p4", "p4", 1); // link to p4 ndn::FibHelper::AddRoute("p1", "/data/p1", "n2", 1); ndn::FibHelper::AddRoute("n2", "/data", "n12", 1); ndn::FibHelper::AddRoute("n12", "/data", "n1", 2); ndn::FibHelper::AddRoute("n1", "/data", "c1", 1); Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Tue Mar 3 16:54:10 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Wed, 4 Mar 2015 09:54:10 +0900 Subject: [ndnSIM] Question about producer. In-Reply-To: References: Message-ID: Does anyone know how to forward data packet from p1? ndn::FibHelper::AddRoute("p1", "/data/p1/1", "n2", 1); ndn::FibHelper::AddRoute("n2", "/data/p1/1", "n12", 1); ndn::FibHelper::AddRoute("n12", "/data/p1/1", "n1", 2); ndn::FibHelper::AddRoute("n1", "/data/p1/1", "c1", 1); I am not sure this way is right. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Tue Mar 3 22:22:05 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Wed, 4 Mar 2015 15:22:05 +0900 Subject: [ndnSIM] About congestion-alt-topo-plugin.cpp example Message-ID: Hello. I don't know why the consumer do not receive data packet from producer in ndn-congestion-alt-topo-plugin.cpp example Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From jpram94 at gmail.com Tue Mar 3 23:13:15 2015 From: jpram94 at gmail.com (Ram Dinesh) Date: Tue, 3 Mar 2015 23:13:15 -0800 Subject: [ndnSIM] Data packet forwarding from intermediate router. Message-ID: Hello, My project is about Mitigating the interest flooding attack. Finally I have to stop the attacker from sending malicious interest packet. For that , I need to send fake data packet to the attacker. Is it possible to send fake data packet from intermediate routers.? if so.. how.? regards, Ram Dinesh -------------- next part -------------- An HTML attachment was scrubbed... URL: From lanwang at memphis.edu Wed Mar 4 07:13:11 2015 From: lanwang at memphis.edu (Lan Wang (lanwang)) Date: Wed, 4 Mar 2015 15:13:11 +0000 Subject: [ndnSIM] How make a inter-domain network by ndnSIM? In-Reply-To: <1378392944.2125914.1425337621211.JavaMail.yahoo@mail.yahoo.com> References: <35019046-0F0E-45D6-8504-57C2828E0F57@memphis.edu> <1378392944.2125914.1425337621211.JavaMail.yahoo@mail.yahoo.com> Message-ID: <34D0E28E-9E4B-4FFD-B49F-D14F54889D82@memphis.edu> I'm not aware of any existing guidelines. But I guess you can run a proxy on one router that converts the HTTP requests to NDN Interests and another proxy on the other router that converts HTTP responses to NDN Data packets. I don't understand why you want to do this kind of translations though. It may be much easier to just rewrite the existing applications to NDN applications. Lan On Mar 2, 2015, at 5:07 PM, amin1431 > wrote: Hi, Yes, I want to connect two different traffics through the NDN. You said for some applications such as http, it is possible. Is there guideline? Sincerely, Armin ________________________________ From: Lan Wang (lanwang) > To: amin1431 > Cc: Spyridon (Spyros) Mastorakis >; Ndnsim > Sent: Monday, March 2, 2015 11:47 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Armin, Do you mean that the traffic in the clusters is not NDN traffic but you want to convert the traffic to NDN at the border routers? I'm not sure how to do this. For some applications (e.g., http like traffic), this may be possible. But it's application specific, not something you can do regardless of the application. Lan On Mar 2, 2015, at 11:45 AM, amin1431 > wrote: Hi Lan, Thank you for your guidance. I want actually connect two clusters (such as two smart grids or two vehicle networks) through NDN. I am trying to get the fundamental understanding to how it is possible to connect two clusters via NDN. for instance, how change the traffic from clusters to the Interest/Data and in reverse. If there is any useful material, I am very thankful to share with me. Sincerely, Armin ________________________________ From: Lan Wang (lanwang) > To: Spyridon (Spyros) Mastorakis > Cc: amin1431 >; Ndnsim > Sent: Monday, March 2, 2015 5:56 PM Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? Not sure exactly what Armin is looking for. The following tutorial is not for ndnSIM, but how to connect the nfd instances running on different machines (not directly connected): http://yoursunny.com/t/2014/nfd-connect/ Lan On Mar 2, 2015, at 10:47 AM, "Spyridon (Spyros) Mastorakis" > wrote: > Hello Armin, > > to the best of my knowledge, we have not published such a tutorial. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > >> On Mar 1, 2015, at 4:51 PM, amin1431 > wrote: >> >> Hi Spyros, >> Thank you for your help. >> Is there any guideline/tutorial to how we can run NDN (or other infrastructures for example) on top of a network? >> >> >> >> Sincerely, >> Armin >> >> ----------------------------------------------------------------------------------------------------------- >> Subject: Re: [ndnSIM] How make a inter-domain network by ndnSIM? >> Date: Wed, 4 Feb 2015 18:06:38 -0800 >> From: Spyridon (Spyros) Mastorakis > >> To: amin1431 > >> CC: Ndnsim > >> >> Hello, >> >> NDN can run on top of every protocol and ndnSIM implements this primitive. >> Therefore, you have to use the implementation of the protocol that you want in your scenario and then run NDN on top of it. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: >> http://cs.ucla.edu/~mastorakis/ >> >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >>> On Feb 3, 2015, at 6:21 AM, amin1431 >> > >> wrote: >>> >>> Hi friends :-) >>> I am going to connect two different domains, such as NDN and Avionics Full-Duplex Switched Ethernet (AFDX). I mean, an inter-domain connection between NDN and AFDX. >>> >>> Could you please give me some hints, how does it possible? >>> >>> >>> Sincerely, >>> Armin >>> >> >> >> _______________________________________________ >> ndnSIM mailing list >> >> ndnSIM at lists.cs.ucla.edu >> http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim >> >> >> >> > > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From wadekahn at clarkson.EDU Wed Mar 4 08:17:46 2015 From: wadekahn at clarkson.EDU (Hitesh Wadekar) Date: Wed, 4 Mar 2015 11:17:46 -0500 Subject: [ndnSIM] Mechanism to check neighbor (link) alive or not Message-ID: Hey Guys, I just gone through NLSR paper, where they have mentioned that they are sending status (echo) messages to neighbor to check whether its alive or not (basically for tracking) which is a kind pf ping utility. I wonder is there any other mechanism in NDN protocol stack that we can check or verify weather connected link or neighbor node is reachable or alive or not? Thank you guys. Cheer, Hitesh -------------- next part -------------- An HTML attachment was scrubbed... URL: From lanwang at memphis.edu Wed Mar 4 08:37:47 2015 From: lanwang at memphis.edu (Lan Wang (lanwang)) Date: Wed, 4 Mar 2015 16:37:47 +0000 Subject: [ndnSIM] Mechanism to check neighbor (link) alive or not In-Reply-To: References: Message-ID: <4DB926F4-480D-4E2A-845A-3E45A0FCA79A@memphis.edu> Are you talking about the Hello messages (we call it INFO message in the paper)? This is something NLSR sends, since it's used to detect whether NLSR is alive or not. So you can't use something else to replace it. For example, if there's link layer mechanism to detect link failures, that will help trigger routing calculations faster in this case. But the link layer failure detection is not going to detect NLSR failure. That said, NLSR does receive notifications from the local NFD when a face goes down or up. Lan On Mar 4, 2015, at 10:17 AM, Hitesh Wadekar > wrote: Hey Guys, I just gone through NLSR paper, where they have mentioned that they are sending status (echo) messages to neighbor to check whether its alive or not (basically for tracking) which is a kind pf ping utility. I wonder is there any other mechanism in NDN protocol stack that we can check or verify weather connected link or neighbor node is reachable or alive or not? Thank you guys. Cheer, Hitesh _______________________________________________ ndnSIM mailing list ndnSIM at lists.cs.ucla.edu http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: From spiros.mastorakis at gmail.com Wed Mar 4 08:48:41 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Wed, 4 Mar 2015 08:48:41 -0800 Subject: [ndnSIM] Data packet forwarding from intermediate router. In-Reply-To: References: Message-ID: Hello, you probably need to code an application and install it on the intermediate routers. This application can send the data packets that you want. However, if your data packets are unsolicited, NFD will drop them for security reasons. In this way, if a ?fake? packet satisfies a PIT entry, the original packet, which will be received later, would be considered as unsolicited and get dropped. If you want to forward both the ?fake? and the original packet, you will have to take care of maintaining the PIT entry after the arrival of the ?fake? packet. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 3, 2015, at 11:13 PM, Ram Dinesh wrote: > > Hello, > > My project is about Mitigating the interest flooding attack. Finally I have to stop the attacker from sending malicious interest packet. For that , I need to send fake data packet to the attacker. Is it possible to send fake data packet from intermediate routers.? if so.. how.? > > regards, > > Ram Dinesh > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From wadekahn at clarkson.EDU Wed Mar 4 09:03:05 2015 From: wadekahn at clarkson.EDU (Hitesh Wadekar) Date: Wed, 4 Mar 2015 12:03:05 -0500 Subject: [ndnSIM] Mechanism to check neighbor (link) alive or not In-Reply-To: <4DB926F4-480D-4E2A-845A-3E45A0FCA79A@memphis.edu> References: <4DB926F4-480D-4E2A-845A-3E45A0FCA79A@memphis.edu> Message-ID: HI Prof. Wang, Yes, it is "Hello" messages. It was typo ( NLSR sends "info" interest), I apologies for it. I completely agree on it. Now I got to know NLSR failure and detections. Yes, It will help to trigger routing calculation faster, Thank you for clarifying :) Regards, Hitesh On Wed, Mar 4, 2015 at 11:37 AM, Lan Wang (lanwang) wrote: > Are you talking about the Hello messages (we call it INFO message in the > paper)? This is something NLSR sends, since it's used to detect whether > NLSR is alive or not. So you can't use something else to replace it. For > example, if there's link layer mechanism to detect link failures, that will > help trigger routing calculations faster in this case. But the link layer > failure detection is not going to detect NLSR failure. > > That said, NLSR does receive notifications from the local NFD when a > face goes down or up. > > Lan > > On Mar 4, 2015, at 10:17 AM, Hitesh Wadekar > wrote: > > Hey Guys, > > I just gone through NLSR paper, where they have mentioned that they are > sending status (echo) messages to neighbor to check whether its alive or > not (basically for tracking) which is a kind pf ping utility. > > I wonder is there any other mechanism in NDN protocol stack that we can > check or verify weather connected link or neighbor node is reachable or > alive or not? > > Thank you guys. > > Cheer, > Hitesh > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexander.afanasyev at ucla.edu Wed Mar 4 09:40:32 2015 From: alexander.afanasyev at ucla.edu (Alex Afanasyev) Date: Wed, 4 Mar 2015 09:40:32 -0800 Subject: [ndnSIM] Question about producer. In-Reply-To: References: Message-ID: You cannot just forward data packet in NDN. There has to be interest received by p1 producer for the data packet to travel back to the consumer. FibHelper::AddRoute provide configuration for the interest paths. -- Alex > On Mar 3, 2015, at 4:54 PM, ??? wrote: > > Does anyone know how to forward data packet from p1? > > ndn::FibHelper::AddRoute("p1", "/data/p1/1", "n2", 1); > ndn::FibHelper::AddRoute("n2", "/data/p1/1", "n12", 1); > ndn::FibHelper::AddRoute("n12", "/data/p1/1", "n1", 2); > ndn::FibHelper::AddRoute("n1", "/data/p1/1", "c1", 1); > > I am not sure this way is right. > > Wonjun, Choi -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 841 bytes Desc: Message signed with OpenPGP using GPGMail URL: From wonjunchoi001 at gmail.com Wed Mar 4 16:17:15 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Thu, 5 Mar 2015 09:17:15 +0900 Subject: [ndnSIM] Question about producer. In-Reply-To: References: Message-ID: Thanks for your reply. I ran ndn-congestion-alt-topo-plugin example but the consumer couldn't get data packet from p1 that is why I am asking you. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Thu Mar 5 23:18:40 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Fri, 6 Mar 2015 16:18:40 +0900 Subject: [ndnSIM] Question about producer. In-Reply-To: References: Message-ID: I am still digging to find the reason why consumer couldn't receive data packet from p1. If someone knows about this. please let me know. ndn::FibHelper::AddRoute is only related to interest packet not data packet? Best regards. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Thu Mar 5 23:21:31 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Fri, 6 Mar 2015 16:21:31 +0900 Subject: [ndnSIM] Question about producer. In-Reply-To: References: Message-ID: > FibHelper::AddRoute provide configuration for the interest paths. What is for the data paths? Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From kanin at inl.ics.keio.ac.jp Fri Mar 6 01:09:45 2015 From: kanin at inl.ics.keio.ac.jp (Kanin Assantachai) Date: Fri, 06 Mar 2015 18:09:45 +0900 Subject: [ndnSIM] Implementing My Own Cache Policy Message-ID: <54F96ED9.3040603@inl.ics.keio.ac.jp> Hello all, After I have read through documentations of ndnSIM 1.0, especially the source codes of the provided basic cache policies, I have some questions about them as well as questions about trying to create my own cache policy as follows. 1. Referring to the source code of LRU policy (lru-policy.h), what are the roles (functions) of codes from line 39 to line 47? I have crossed check with the code of empty policy (empty-policy.h) and I found that such codes are not necessary for the case of policy that does not implement any algorithm. 2. From line 89 to 96 of the same file (lru-policy.h), the "lookup" function is implemented there. Why doesn't it handle the case of cache miss? Based on the code, it seems to me that this case is not handled at all. 3. It seems to me that the "policy_container" in the code serves as a cache storage (list). In that case, how can I check for the list of available functions I can call? I wonder if there are functions I can use to access specific orders of cached objects in the list. Is there a way to set IDE to automatically display a list of possible related functions? 4. This question is about the policy I would like to implement. How can I make nodes automatically send messages about some of the contents of their stored cache objects to next neighbor nodes? Is it better to implement this detail in the same cache policy file or different files at different levels? Please kindly provide some suggestions. Thank you for your responses in advance. Kanin Assantachai From wadekahn at clarkson.edu Fri Mar 6 12:59:02 2015 From: wadekahn at clarkson.edu (Hitesh Wadekar) Date: Fri, 6 Mar 2015 15:59:02 -0500 Subject: [ndnSIM] Steps to create control commands from application Message-ID: Hey Friends, I am looking for control commands. How can I generate control commands from application? what are the steps for it? any references or links would be helpful. For instance, I want to add or updated 'routes' or nexthops in FIB using control command ("/localhost/nfd/fib") from application. I tried to search in mailing list first, but I wasn't able to find any comments or references on it. Also, I think currently, in NFD, its not possible that remote application can configure destination NFD using control command? Please correct me if I am wrong. Cheer, Hitesh -------------- next part -------------- An HTML attachment was scrubbed... URL: From aloulounarjes at yahoo.fr Thu Mar 12 12:26:32 2015 From: aloulounarjes at yahoo.fr (narjes aloulou) Date: Thu, 12 Mar 2015 19:26:32 +0000 Subject: [ndnSIM] Incoming Interest In-Reply-To: <4A4B37FA-0C6B-48C9-B950-A4ABAB670863@memphis.edu> Message-ID: <1426188392.91855.YahooMailBasic@web172106.mail.ir2.yahoo.com> Hi, I want to send a control message from a node to its neighbor. So, I use an Interest with a new Nack value. My question is how can I identify a message sent by the node's application layer or a message received from its neighbor (in OnInterest method in forwardingstarategy class). I try to use the following code, but it doesn't work. TypeId faceType = inFace->GetInstanceTypeId(); std::string faceType_name = faceType.GetName(); if(faceType_name.compare("ns3::ndn::AppFace") == 0 ) { // from my application } else { // from my neighbor } Can you give me a hint? Best reqards. From trade20122012 at gmail.com Fri Mar 13 09:32:57 2015 From: trade20122012 at gmail.com (Trade) Date: Fri, 13 Mar 2015 09:32:57 -0700 Subject: [ndnSIM] get prefix seqno Message-ID: <55031139.1010308@gmail.com> I need to do with multiple prefixes, but how can I grab the seq no of prefix so that I can work with prefixes as a list ? Thanks. Trade -------------- next part -------------- An HTML attachment was scrubbed... URL: From hishamfarahat at gmail.com Fri Mar 20 21:33:26 2015 From: hishamfarahat at gmail.com (Hisham Farahat) Date: Sat, 21 Mar 2015 00:33:26 -0400 Subject: [ndnSIM] How not to cache specific data packets? Message-ID: Hi, I'm using ndnSim 1.0. Is there a way to disable caching for some data packets? For example, data packets named with /prefix_10 will not be cached while all other data packets will be cached normally. Regards, -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexander.afanasyev at ucla.edu Fri Mar 20 22:59:26 2015 From: alexander.afanasyev at ucla.edu (Alex Afanasyev) Date: Fri, 20 Mar 2015 22:59:26 -0700 Subject: [ndnSIM] How not to cache specific data packets? In-Reply-To: References: Message-ID: <0720B69B-6BB8-42C1-8B5C-68647378DBD8@ucla.edu> Hi Hisham, There is no built-in way to do that, but you can always hack in the replacement policy and add something like in insert method (see model/cs/custom-policies/* as a reference): inline bool insert (typename parent_trie::iterator item) { static const Name& prefix("/prefix_10?); const Name& name = item->payload ()->GetData ()->GetName (); if (prefix.IsPrefixOf(name)) { return false; } ... whatever existing code in insert method } ? Alex > On Mar 20, 2015, at 9:33 PM, Hisham Farahat wrote: > > Hi, > I'm using ndnSim 1.0. > Is there a way to disable caching for some data packets? > For example, data packets named with /prefix_10 will not be cached while all other data packets will be cached normally. > > Regards, > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 841 bytes Desc: Message signed with OpenPGP using GPGMail URL: From alexander.afanasyev at ucla.edu Fri Mar 20 23:01:25 2015 From: alexander.afanasyev at ucla.edu (Alex Afanasyev) Date: Fri, 20 Mar 2015 23:01:25 -0700 Subject: [ndnSIM] get prefix seqno In-Reply-To: <55031139.1010308@gmail.com> References: <55031139.1010308@gmail.com> Message-ID: Hi Trade, I?m not sure I fully understand your question. Assuming sequence number is at the last component of data name, you can do something like (in ndnSIM 2.0, 1.0 will have slightly different syntax) data.getName().get(-1).toSeqNo(); Did you mean something else? ? Alex > On Mar 13, 2015, at 9:32 AM, Trade wrote: > > I need to do with multiple prefixes, but how can I grab the seq no of prefix > so that I can work with prefixes as a list ? > Thanks. > Trade > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 841 bytes Desc: Message signed with OpenPGP using GPGMail URL: From wonjunchoi001 at gmail.com Sun Mar 22 02:53:00 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Sun, 22 Mar 2015 18:53:00 +0900 Subject: [ndnSIM] custom fw strategy Message-ID: Hello. I want to make custom fw strategy I made best-route3.cpp, hpp files in daemon/fw folder I declared like this: const Name BestRouteStrategy3::STRATEGY_NAME("ndn:/localhost/nfd/strategy/best-route3"); I ran this code. ndn::StrategyChoiceHelper::InstallAll("/prefix", "/localhost/nfd/strategy/best-route3"); but it was not working. I don't know why. please let me know. Best regards. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexander.afanasyev at ucla.edu Sun Mar 22 07:33:50 2015 From: alexander.afanasyev at ucla.edu (Alex Afanasyev) Date: Sun, 22 Mar 2015 09:33:50 -0500 Subject: [ndnSIM] custom fw strategy In-Reply-To: References: Message-ID: <22088CCF-DC25-4413-ABCF-7458A2462C3E@ucla.edu> Hi, Before running InstallAll method, you would also need to call StrategyChoiceHelper::Install(someNode, prefix); (this will make sure strategy is available on node ?someNode?). ? Alex > On Mar 22, 2015, at 4:53 AM, ??? wrote: > > Hello. > I want to make custom fw strategy > I made best-route3.cpp, hpp files in daemon/fw folder > > I declared like this: > const Name BestRouteStrategy3::STRATEGY_NAME("ndn:/localhost/nfd/strategy/best-route3"); > > I ran this code. > ndn::StrategyChoiceHelper::InstallAll("/prefix", "/localhost/nfd/strategy/best-route3"); > > but it was not working. I don't know why. please let me know. > > Best regards. > Wonjun, Choi > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 841 bytes Desc: Message signed with OpenPGP using GPGMail URL: From wonjunchoi001 at gmail.com Sun Mar 22 22:51:55 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Mon, 23 Mar 2015 14:51:55 +0900 Subject: [ndnSIM] how can I get received data time? Message-ID: Hello. I declared two variables in forwarder.hpp forwarder.hpp 207 private: 208 time::steady_clock::TimePoint IStime; // Sent Interest time 209 time::steady_clock::TimePoint DRtime; // Received Data time. how can I get those values in best-route-strategy.cpp? I hope someone can give an example. Best regards. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From spiros.mastorakis at gmail.com Sun Mar 22 23:02:57 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Sun, 22 Mar 2015 23:02:57 -0700 Subject: [ndnSIM] how can I get received data time? In-Reply-To: References: Message-ID: Hello, there are some calls of a method ?dispatchToStrategy" in forwarder.cpp. This method will find and call the right forwarding strategy. You can pass your variables as arguments to this method. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 22, 2015, at 10:51 PM, ??? wrote: > > Hello. > > I declared two variables in forwarder.hpp > > forwarder.hpp > 207 private: > 208 time::steady_clock::TimePoint IStime; // Sent Interest time > 209 time::steady_clock::TimePoint DRtime; // Received Data time. > > how can I get those values in best-route-strategy.cpp? > I hope someone can give an example. > > Best regards. > Wonjun, Choi > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From christian.kreuzberger at itec.aau.at Mon Mar 23 07:23:54 2015 From: christian.kreuzberger at itec.aau.at (Christian Kreuzberger) Date: Mon, 23 Mar 2015 15:23:54 +0100 Subject: [ndnSIM] scenario template in ndnSim 2.0 Message-ID: <006401d06574$fe5cca10$fb165e30$@itec.aau.at> Hi, not sure if it is on purpose, but if I do git clone https://github.com/cawka/ns-3-dev-ndnSIM.git ns-3 I still get the version without the changes for the ndnSIM 2.0 scenario template. I had to manually switch to the ndnSIM-v2 branch, else I would get the error message that the header file "common.hpp" could not be found. Best regards, Christian -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Tue Mar 24 04:48:17 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Tue, 24 Mar 2015 20:48:17 +0900 Subject: [ndnSIM] Converting from nanoseconds to seconds Message-ID: Hello. How can I convert from nanoseconds to seconds? time::nanoseconds time= time::steady_clock::now() time.getseconds() is not working. Best regards. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From alexander.afanasyev at ucla.edu Tue Mar 24 08:27:10 2015 From: alexander.afanasyev at ucla.edu (Alex Afanasyev) Date: Tue, 24 Mar 2015 10:27:10 -0500 Subject: [ndnSIM] Converting from nanoseconds to seconds In-Reply-To: References: Message-ID: <859C5EBC-56E4-4E66-8431-CAA85B2D37FC@ucla.edu> Hi! Given nanoseconds-to-seconds conversion includes information truncation, you would need to explicitly specify this intention using duration_cast method (you can check more about duration_cast in Boost.Chrono documentation http://www.boost.org/doc/libs/1_57_0/doc/html/chrono/users_guide.html or in std::chrono docs http://en.cppreference.com/w/cpp/chrono/duration/duration_cast ): There is also a small problem with your example. time::steady_clock::now() returns a so called ?time point? (time duration since some fixed epoch). More correct example would look similar to: ::ndn::time::steady_clock::TimePoint now = ::ndn::time::steady_clock::now(); ::ndn::time::seconds sec = ::ndn::time::duration_cast<::ndn::time::seconds>(now.time_since_epoch()); ? Alex > On Mar 24, 2015, at 6:48 AM, ??? wrote: > > Hello. > > How can I convert from nanoseconds to seconds? > time::nanoseconds time= time::steady_clock::now() > time.getseconds() is not working. > > Best regards. > Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 841 bytes Desc: Message signed with OpenPGP using GPGMail URL: From anilj.mailing at gmail.com Tue Mar 24 16:33:56 2015 From: anilj.mailing at gmail.com (Anil Jangam) Date: Tue, 24 Mar 2015 16:33:56 -0700 Subject: [ndnSIM] How to select multiple faces to send Interest/Data? Message-ID: https://github.com/named-data/ndnSIM/blob/master/examples/ndn-custom-apps/custom-app.cpp In this custom-app.cc code (line: 88), I think it is sending the Interest on face created by the application (ndn-app). However, how the scenario is handled where there are more than one faces on given node? How a particular face is selected to send the Interest on? In my simulation setup, I have multiple faces on given node. When Interest/Data is received, a respective handler is called (line: 93 or 116). I believe this handler is called for all Interests/Data messages received from all interface, correct? It is the application, based on the Interest name, carries out different processing as per the application context. Am I correct? I saw in one of the application code, an ExpressInterest() API, where one can specify the exact call back to handle Data. However in the above code, this looks like abstracted out. Is it still possible to use the ExpressInterest API? To put it other words, I did not understand how Interest is actually being sent out to the face. How does this works in the above sample code? /anil. -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Tue Mar 24 19:37:56 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Wed, 25 Mar 2015 11:37:56 +0900 Subject: [ndnSIM] no member named 'GetNode' in 'nfd::Face' Message-ID: Hello. I printed like this but it occurs error. inFace->GetNode()->GetId() in forwader.cpp file. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From spiros.mastorakis at gmail.com Tue Mar 24 20:29:21 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Tue, 24 Mar 2015 20:29:21 -0700 Subject: [ndnSIM] How to select multiple faces to send Interest/Data? In-Reply-To: References: Message-ID: <84FBBDB9-085E-4A7F-BD43-BAA7C648414E@gmail.com> Hello Anil, I will try to answer your questions: First of all, the application face is different than the NetDevice Face, which is used for the connection among the simulated nodes. Both of them are realizations of the NFD Face class. The NFD Face class is the base class, which is used during the Interest/Data processing from NFD. For further details, you can take a look at our Technical Report: http://named-data.net/publications/techreports/ndn-0028-1-ndnsim-v2/ > In this custom-app.cc code (line: 88), I think it is sending the Interest on face created by the application (ndn-app). However, how the scenario is handled where there are more than one faces on given node? How a particular face is selected to send the Interest on? In my simulation setup, I have multiple faces on given node. Each application, that you will define in a simulation scenario using the AppHelper, will have a face associated with it. In this way, an Interest packet will be sent out of this application face to NFD for further processing and then to other nodes using the netDeviceFace. I do not think that there are any reasons to have multiple faces for the same application instance on a single node. However, you can install multiple application instances on the same node. > When Interest/Data is received, a respective handler is called (line: 93 or 116). I believe this handler is called for all Interests/Data messages received from all interface, correct? It is the application, based on the Interest name, carries out different processing as per the application context. Am I correct? What do you mean by ?all interfaces?? Application faces, NFD faces, NetDeviceFaces? > I saw in one of the application code, an ExpressInterest() API, where one can specify the exact call back to handle Data. However in the above code, this looks like abstracted out. Is it still possible to use the ExpressInterest API? To put it other words, I did not understand how Interest is actually being sent out to the face. How does this works in the above sample code? An Interest starting from the application face will be send to NFD for further processing. If there is a FIB entry for the expressed name prefix, the interest will be sent to another node. Hope that this helps. -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA From anilj.mailing at gmail.com Tue Mar 24 22:29:19 2015 From: anilj.mailing at gmail.com (Anil Jangam) Date: Tue, 24 Mar 2015 22:29:19 -0700 Subject: [ndnSIM] How to select multiple faces to send Interest/Data? In-Reply-To: <84FBBDB9-085E-4A7F-BD43-BAA7C648414E@gmail.com> References: <84FBBDB9-085E-4A7F-BD43-BAA7C648414E@gmail.com> Message-ID: Hi Spyridon. Thank you for your comments. They are elaborate! I have added few further notes. I am writing a simulation for a protocol where multiple nodes are involved. In my topology, a given node may be connected to multiple other nodes. I have written an application (derives from "ns3::ndn::App" class), which is installed on node. Right now, I am more thinking from this application perspective, hence my earlier email. I will read the technical report. /anil. On Tue, Mar 24, 2015 at 8:29 PM, Spyridon (Spyros) Mastorakis < spiros.mastorakis at gmail.com> wrote: > Hello Anil, > > I will try to answer your questions: > > First of all, the application face is different than the NetDevice Face, > which is used for the connection among the simulated nodes. Both of them > are realizations of the NFD Face class. The NFD Face class is the base > class, which is used during the Interest/Data processing from NFD. For > further details, you can take a look at our Technical Report: > > http://named-data.net/publications/techreports/ndn-0028-1-ndnsim-v2/ > > > In this custom-app.cc code (line: 88), I think it is sending the > Interest on face created by the application (ndn-app). However, how the > scenario is handled where there are more than one faces on given node? How > a particular face is selected to send the Interest on? In my simulation > setup, I have multiple faces on given node. > > Each application, that you will define in a simulation scenario using the > AppHelper, will have a face associated with it. In this way, an Interest > packet will be sent out of this application face to NFD for further > processing and then to other nodes using the netDeviceFace. I do not think > that there are any reasons to have multiple faces for the same application > instance on a single node. However, you can install multiple application > instances on the same node. > [anil] Ok. "more than one faces on given node" - I was talking about when a node is connected to more than one remote nodes. So if I understood you correctly, there will be multiple netDeviceFace(s) depending on how node is connected to other nodes. I agree that there is no need to have more than one application faces between application and NFD. However, if I have to send an Interest to specific node, can't I use a specific netDeviceFace instead of depending on NFD and FIB entries? Is that possible? > > When Interest/Data is received, a respective handler is called (line: 93 > or 116). I believe this handler is called for all Interests/Data messages > received from all interface, correct? It is the application, based on the > Interest name, carries out different processing as per the application > context. Am I correct? > > What do you mean by ?all interfaces?? Application faces, NFD faces, > NetDeviceFaces? > [anil] I meant the Interest/Data messages received from remote devices. So it is netDeviceFace. But now, I see that these handlers (OnInterest, and OnData) are application level handlers (point of entry), and the lower layer (e.g. NFD) is going to post Interest/Data messages to the application over application face. Application then handle/process these Interest/Data depending on the "name"space, correct? > > I saw in one of the application code, an ExpressInterest() API, where > one can specify the exact call back to handle Data. However in the above > code, this looks like abstracted out. Is it still possible to use the > ExpressInterest API? To put it other words, I did not understand how > Interest is actually being sent out to the face. How does this works in the > above sample code? > > An Interest starting from the application face will be send to NFD for > further processing. If there is a FIB entry for the expressed name prefix, > the interest will be sent to another node. > [anill] I was more concerned about the Ingress Interest/Data message. So to extend my above comment, when application receives Interest/Data message, the OnInterest/OnData is called by the simulator framework respectively, and/or we can specify the function to be called back - on receiving Interest/Data - by using ExpressInterest() API. Is this correct? > Hope that this helps. > > -- > Spyridon (Spyros) Mastorakis > Personal Website: http://cs.ucla.edu/~mastorakis/ > Internet Research Laboratory > PhD Computer Science > UCLA > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From wonjunchoi001 at gmail.com Wed Mar 25 01:17:05 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Wed, 25 Mar 2015 17:17:05 +0900 Subject: [ndnSIM] no member named 'GetNode' in 'nfd::Face' In-Reply-To: References: Message-ID: Please let me know how to print node id in forwarder.cpp? Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From kanin at inl.ics.keio.ac.jp Wed Mar 25 10:13:57 2015 From: kanin at inl.ics.keio.ac.jp (Kanin Assantachai) Date: Thu, 26 Mar 2015 02:13:57 +0900 Subject: [ndnSIM] Implementing My Own Cache Management Policy Message-ID: <5512ECD5.5020004@inl.ics.keio.ac.jp> Hello all, I am currently using ndnSIM 1.0, and I have been trying to implement my own cache management policy for a while already. However, I have quite a lot of questions to ask as follows. 1. Referring to the source code of LRU policy (lru-policy.h), what are the roles (functions) of codes from line 39 to line 47? I have crossed check with the code of empty policy (empty-policy.h) and I found that such codes are not necessary for the case of policy that does not implement any algorithm. 2. From line 89 to 96 of the same file (lru-policy.h), the "lookup" function is implemented there. Why doesn't it handle the case of cache miss? Based on the code, it seems to me that this case is not handled at all. 3. It seems to me that the "policy_container" in the code serves as a cache storage (list). I wonder if there are functions I can use to access specific orders of cached objects in the list at a particular time. Let's say I would like to get the first 5 contents which are stored in the cache storage. How can I possibly do that? 4. This question is about the policy I would like to implement. How can I make nodes automatically send messages about some of the contents of their stored cache objects to next neighbor nodes? Is it better to implement this detail in the same cache policy file that I am going to create, or in different files at different levels? Please kindly provide some suggestions. 5. When I simulated the example file "ndn-simple-with-cs-lfu.cc" using the visualizer, I discovered that the request from consumer node still reaches the producer node even after the intermediate router has received the requested content from producer at least once. (The second request from consumer node is sent before the content in the cache of the router expires.) If cache is enabled at the router, why does the later requests (after the first one) still reaches the producer node? Why doesn't the router serve the request with its cache content? Thank you for your responses in advance. Kanin Assantachai From anilj.mailing at gmail.com Wed Mar 25 10:31:36 2015 From: anilj.mailing at gmail.com (Anil Jangam) Date: Wed, 25 Mar 2015 10:31:36 -0700 Subject: [ndnSIM] How to select multiple faces to send Interest/Data? In-Reply-To: References: <84FBBDB9-085E-4A7F-BD43-BAA7C648414E@gmail.com> Message-ID: Hi Spyros, Kindly please note that I'm using ndnSIM 1.0. /anil On Mar 24, 2015 10:29 PM, "Anil Jangam" wrote: > Hi Spyridon. > > Thank you for your comments. They are elaborate! I have added few further > notes. > > I am writing a simulation for a protocol where multiple nodes are > involved. In my topology, a given node may be connected to multiple other > nodes. I have written an application (derives from "ns3::ndn::App" class), > which is installed on node. Right now, I am more thinking from this > application perspective, hence my earlier email. > > I will read the technical report. > > /anil. > > > On Tue, Mar 24, 2015 at 8:29 PM, Spyridon (Spyros) Mastorakis < > spiros.mastorakis at gmail.com> wrote: > >> Hello Anil, >> >> I will try to answer your questions: >> >> First of all, the application face is different than the NetDevice Face, >> which is used for the connection among the simulated nodes. Both of them >> are realizations of the NFD Face class. The NFD Face class is the base >> class, which is used during the Interest/Data processing from NFD. For >> further details, you can take a look at our Technical Report: >> >> http://named-data.net/publications/techreports/ndn-0028-1-ndnsim-v2/ >> >> > In this custom-app.cc code (line: 88), I think it is sending the >> Interest on face created by the application (ndn-app). However, how the >> scenario is handled where there are more than one faces on given node? How >> a particular face is selected to send the Interest on? In my simulation >> setup, I have multiple faces on given node. >> >> Each application, that you will define in a simulation scenario using the >> AppHelper, will have a face associated with it. In this way, an Interest >> packet will be sent out of this application face to NFD for further >> processing and then to other nodes using the netDeviceFace. I do not think >> that there are any reasons to have multiple faces for the same application >> instance on a single node. However, you can install multiple application >> instances on the same node. >> > > [anil] Ok. "more than one faces on given node" - I was talking about when > a node is connected to more than one remote nodes. So if I understood you > correctly, there will be multiple netDeviceFace(s) depending on how node is > connected to other nodes. > I agree that there is no need to have more than one application faces > between application and NFD. However, if I have to send an Interest to > specific node, can't I use a specific netDeviceFace instead of depending on > NFD and FIB entries? Is that possible? > > >> > When Interest/Data is received, a respective handler is called (line: >> 93 or 116). I believe this handler is called for all Interests/Data >> messages received from all interface, correct? It is the application, based >> on the Interest name, carries out different processing as per the >> application context. Am I correct? >> >> What do you mean by ?all interfaces?? Application faces, NFD faces, >> NetDeviceFaces? >> > > [anil] I meant the Interest/Data messages received from remote devices. So > it is netDeviceFace. But now, I see that these handlers (OnInterest, and > OnData) are application level handlers (point of entry), and the lower > layer (e.g. NFD) is going to post Interest/Data messages to the application > over application face. Application then handle/process these Interest/Data > depending on the "name"space, correct? > > >> > I saw in one of the application code, an ExpressInterest() API, where >> one can specify the exact call back to handle Data. However in the above >> code, this looks like abstracted out. Is it still possible to use the >> ExpressInterest API? To put it other words, I did not understand how >> Interest is actually being sent out to the face. How does this works in the >> above sample code? >> >> An Interest starting from the application face will be send to NFD for >> further processing. If there is a FIB entry for the expressed name prefix, >> the interest will be sent to another node. >> > > [anill] I was more concerned about the Ingress Interest/Data message. So > to extend my above comment, when application receives Interest/Data > message, the OnInterest/OnData is called by the simulator framework > respectively, and/or we can specify the function to be called back - on > receiving Interest/Data - by using ExpressInterest() API. Is this correct? > > >> Hope that this helps. >> >> -- >> Spyridon (Spyros) Mastorakis >> Personal Website: http://cs.ucla.edu/~mastorakis/ >> Internet Research Laboratory >> PhD Computer Science >> UCLA >> >> >> >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From ioannoa at scss.tcd.ie Wed Mar 25 11:01:16 2015 From: ioannoa at scss.tcd.ie (ioannoa) Date: Wed, 25 Mar 2015 18:01:16 +0000 Subject: [ndnSIM] Introduce delay Message-ID: <9a52d4557794c7027525a1386def4312@scss.tcd.ie> Hello everyone, I am looking on how I can calculate the cost that a PIT lookup would introduce. I have found a solution where a Schedule Event may be used for it but I want this to be called in e.g. my OnInterest or OnData function. However, I haven't found a way that I can force the function to "wait" until the execution of this event occurs. In this relevant post http://www.lists.cs.ucla.edu/pipermail/ndnsim/2013-May/000542.html I found that additional modules are necessary so as to implement something like this. Could someone please advice me on how to proceed? Thanks in advance, A. From wonjunchoi001 at gmail.com Wed Mar 25 17:58:10 2015 From: wonjunchoi001 at gmail.com (=?UTF-8?B?7LWc7JuQ7KSA?=) Date: Thu, 26 Mar 2015 09:58:10 +0900 Subject: [ndnSIM] no member named 'GetNode' in 'nfd::Face' In-Reply-To: References: Message-ID: I am still digging about how to get the node id in forwarder class. If someone have some hint. please let me know. Thanks. Wonjun, Choi -------------- next part -------------- An HTML attachment was scrubbed... URL: From spiros.mastorakis at gmail.com Wed Mar 25 18:12:13 2015 From: spiros.mastorakis at gmail.com (Spyridon (Spyros) Mastorakis) Date: Wed, 25 Mar 2015 18:12:13 -0700 Subject: [ndnSIM] no member named 'GetNode' in 'nfd::Face' In-Reply-To: References: Message-ID: Hello, the most convenient way that I can think of right now is to define get/set methods in the forwarder script in order to get/set a ns3::Ptr to the node as a private variable. You will also need to modify a little bit the L3 Protocol class. I have done something like this here: https://github.com/spirosmastorakis/ndnSIM-development/blob/master/model/ndn-l3-protocol.cc#L97 https://github.com/spirosmastorakis/ndnSIM-development/blob/master/NFD/daemon/fw/forwarder.cpp#L59 -- Spyridon (Spyros) Mastorakis Personal Website: http://cs.ucla.edu/~mastorakis/ Internet Research Laboratory PhD Computer Science UCLA > On Mar 25, 2015, at 5:58 PM, ??? wrote: > > I am still digging about how to get the node id in forwarder class. > If someone have some hint. please let me know. > Thanks. > > Wonjun, Choi > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From kanin at inl.ics.keio.ac.jp Thu Mar 26 19:20:14 2015 From: kanin at inl.ics.keio.ac.jp (Kanin Assantachai) Date: Fri, 27 Mar 2015 11:20:14 +0900 Subject: [ndnSIM] (Please help) Implementing Customized Cache Management Policy In-Reply-To: <5512ECD5.5020004@inl.ics.keio.ac.jp> References: <5512ECD5.5020004@inl.ics.keio.ac.jp> Message-ID: <5514BE5E.1020503@inl.ics.keio.ac.jp> Hello Alex and Spyridon, It's been more than a day since I sent an email about this set of questions already, but I have not received any response yet. I understand that I have asked quite a lot of questions, so it may take a while to completely answer all of them. May I ask if you are working on replying to my questions? Have I done anything wrong in my previous email? I truly need help from you. Kanin Assantachai On 26-Mar-15 2:13 AM, Kanin Assantachai wrote: > Hello all, > > I am currently using ndnSIM 1.0, and I have been trying to implement > my own cache management policy for a while already. However, I have > quite a lot of questions to ask as follows. > > 1. Referring to the source code of LRU policy (lru-policy.h), what > are the roles (functions) of codes from line 39 to line 47? I have > crossed check with the code of empty policy (empty-policy.h) and I > found that such codes are not necessary for the case of policy that > does not implement any algorithm. > > 2. From line 89 to 96 of the same file (lru-policy.h), the "lookup" > function is implemented there. Why doesn't it handle the case of > cache miss? Based on the code, it seems to me that this case is not > handled at all. > > 3. It seems to me that the "policy_container" in the code serves as a > cache storage (list). I wonder if there are functions I can use to > access specific orders of cached objects in the list at a particular > time. Let's say I would like to get the first 5 contents which are > stored in the cache storage. How can I possibly do that? > > 4. This question is about the policy I would like to implement. How > can I make nodes automatically send messages about some of the > contents of their stored cache objects to next neighbor nodes? Is it > better to implement this detail in the same cache policy file that I > am going to create, or in different files at different levels? Please > kindly provide some suggestions. > > 5. When I simulated the example file "ndn-simple-with-cs-lfu.cc" using > the visualizer, I discovered that the request from consumer node still > reaches the producer node even after the intermediate router has > received the requested content from producer at least once. (The > second request from consumer node is sent before the content in the > cache of the router expires.) If cache is enabled at the router, why > does the later requests (after the first one) still reaches the > producer node? Why doesn't the router serve the request with its cache > content? > > Thank you for your responses in advance. > > Kanin Assantachai > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim From narges.mehran at gmail.com Sat Mar 28 07:51:54 2015 From: narges.mehran at gmail.com (Narges Mehran) Date: Sat, 28 Mar 2015 19:21:54 +0430 Subject: [ndnSIM] problem in including path stretch factor! Message-ID: Hello everyone, I am currently using ndnSIM 1.0, and I have been trying to evaluate by a metric like path stretch factor, but while including the hearer "weights-path-stretch-tag.h", I get the {fatal error: weights-path-stretch-tag.h: No such file or directory}! When I searched, I understood that this header file is in a folder named: tracers-broken! Is there any problem with these or my problem is somewhere else? Is there any sample? Thanks in advance -------------- next part -------------- An HTML attachment was scrubbed... URL: From kanin at inl.ics.keio.ac.jp Tue Mar 31 00:15:17 2015 From: kanin at inl.ics.keio.ac.jp (Kanin Assantachai) Date: Tue, 31 Mar 2015 16:15:17 +0900 Subject: [ndnSIM] Implementing My Own Cache Management Policy In-Reply-To: <5512ECD5.5020004@inl.ics.keio.ac.jp> References: <5512ECD5.5020004@inl.ics.keio.ac.jp> Message-ID: <551A4985.2030400@inl.ics.keio.ac.jp> Hello Alex and Spyridon, Since it has been about 5 days already, I would like to resend this email again in case you did not notice the email I sent last Thursday. I truly need your help. Thank you. Kanin Assantachai On 26-Mar-15 2:13 AM, Kanin Assantachai wrote: > Hello all, > > I am currently using ndnSIM 1.0, and I have been trying to implement > my own cache management policy for a while already. However, I have > quite a lot of questions to ask as follows. > > 1. Referring to the source code of LRU policy (lru-policy.h), what > are the roles (functions) of codes from line 39 to line 47? I have > crossed check with the code of empty policy (empty-policy.h) and I > found that such codes are not necessary for the case of policy that > does not implement any algorithm. > > 2. From line 89 to 96 of the same file (lru-policy.h), the "lookup" > function is implemented there. Why doesn't it handle the case of > cache miss? Based on the code, it seems to me that this case is not > handled at all. > > 3. It seems to me that the "policy_container" in the code serves as a > cache storage (list). I wonder if there are functions I can use to > access specific orders of cached objects in the list at a particular > time. Let's say I would like to get the first 5 contents which are > stored in the cache storage. How can I possibly do that? > > 4. This question is about the policy I would like to implement. How > can I make nodes automatically send messages about some of the > contents of their stored cache objects to next neighbor nodes? Is it > better to implement this detail in the same cache policy file that I > am going to create, or in different files at different levels? Please > kindly provide some suggestions. > > 5. When I simulated the example file "ndn-simple-with-cs-lfu.cc" using > the visualizer, I discovered that the request from consumer node still > reaches the producer node even after the intermediate router has > received the requested content from producer at least once. (The > second request from consumer node is sent before the content in the > cache of the router expires.) If cache is enabled at the router, why > does the later requests (after the first one) still reaches the > producer node? Why doesn't the router serve the request with its cache > content? > > Thank you for your responses in advance. > > Kanin Assantachai > > _______________________________________________ > ndnSIM mailing list > ndnSIM at lists.cs.ucla.edu > http://www.lists.cs.ucla.edu/mailman/listinfo/ndnsim