1 /*! \page platform Describing the virtual platform
5 In order to run any simulation, SimGrid must be provided with three things:
6 something to run (i.e., your code), a description of the platform on which you want to simulate your application, and
7 information about the deployment of the application: Which process should be executed onto which processor/core?
9 For the last two items, there are essentially three possible ways you can provide
10 this information as an input:
11 \li You can program, if you're using MSG, some of the platform and
12 deployment functions. If you choose to follow this approach, check the dedicated documentation
13 (\ref msg_simulation).
14 \li You can use two XML files: one for the platform description and the other for the deployment.
15 \li You can program the description of your platform in Lua format.
17 For more information on SimGrid's deployment features, please refer to the \ref deployment section.
19 The platform description may be intricate. This documentation is all about how to write this file. First, the basic
20 concepts are introduced. Then, advanced options are explained. Finally, some hints and tips on how to write a better
21 platform description are given.
23 \section pf_overview Some words about XML and DTD
25 We opted for XML not only because it is extensible but also because many tools (and plugins for existing tools) are
26 available that facilitate editing and validating XML files. Furthermore, libraries that parse XML are often already
27 available and very well tested.
29 The XML checking is done based on the [simgrid.dtd](http://simgrid.gforge.inria.fr/simgrid/simgrid.dtd) Document Type
30 Definition (DTD) file.
32 If you read the DTD, you should notice the following:
33 \li The platform tag has a version attribute. The current version is <b>4</b>. This attribute might be used in the
34 provide backward compatibility.
35 \li The DTD contains definitions for both the platform description and deployment files used by SimGrid.
37 \section pf_basics Basic concepts
39 It is certainly very tempting to defining how the information is
40 routed in the simulated network in a way that is very similar to how
41 it is defined on real network. One would have to define the routing
42 tables of each routers interconnections sub-networks, just like in the
43 real life. Given the daunting amount of configuration required, we
44 could complete the information given by the user with classical
45 protocols such as BGP and RIP. Many network simulator take such
46 configuration as an input, for good reasons.
48 This is not the way it goes in SimGrid: the network routing is defined
49 in a global and compact way instead. This eases the modeling of very
50 large systems, and allows highly optimized datastructures and
51 algorithms in the simulator. The proposed description mechanism is
52 thus much more convinient and efficient. In addition, it is more
53 expressive than the classical solution based on forwarding tables on
56 The price to pay is that this representation of networks is very
57 specific to SimGrid, so you will have to read further to understand
58 it, even if you already know how real networks work.
60 The central notion here are \b Networking \b Zones. NetZones represent
61 network areas in which the routing is done in an homogeneous way.
62 Conceptually, netzones generalize from the ideas of local networks
63 (such as Ethernet switched networks) and Autonomous System. The
64 network as a whole is represented as a single hierarchy of netzones,
65 meaning that every netzone is part of another netzone (but the \c
66 NetRoot, which is the top-level netzone).
68 The main goal of the routing module is to provide a list of links
69 traversed by a given communication and/or a latency to apply. These
70 information are then used by the network model to compute the time
71 that this communication takes. This information is retrieved by three
72 combined algorithms: intra-zone routing, inter-zone routing, and the
75 The <b>intra-zone level</b> is naturally handled by the netzones. Each
76 netzone have to specify the routing algorithm it uses for that.
77 @ref{FullZone} netzones have complete matrix where matrix(a,b)
78 represents the full path (the list of links) between the hosts a and
79 b. @ref{FloydZone} apply the Floyd-Warshall algorithm to compute the
80 paths. @ref{ClusterZone} model classical switched or hub networks,
81 where each component is connected through a private link onto a common
82 backbone. Many other routing algorithms are provided to model the
83 classical needs, but you can naturally define your own routing if the
84 provided ones do not fit your needs.
86 The <b>inter-zone algorithm</b> is used when the communication
87 traverses more than one zone. The overall path goes from the source up
88 in the netzones' tree, until the first common ancestor zone, and moves
89 down to the destination. It crawls the differing netzones on its path
90 according to the user-defined inter-zone routes, moving from gateway
93 You can also use the <b>bypass mechanism</b> to specify manually some
94 shortcuts that directly provide the list of links interconnecting two
99 <AS id="netzone0" routing="Full">
103 Other supported values for the routing attribute can be found below, Section
106 There is also the ``<route>`` tag; this tag takes two attributes,
107 ``src`` (source) and ``dst`` (destination). Both source and
108 destination must be valid identifiers for routers (these will be
109 introduced later). Contained by the ``<route>`` are network links;
110 these links must be used in order to communicate from the source to
111 the destination specified in the tag. Hence, a route merely describes
112 how to reach a router from another router.
115 More information and (code-)examples can be found in Section \ref pf_rm.
117 A netzone can also contain itself one or more netzone; this allows you to model
118 the hierarchy of your platform.
120 ### Within each AS, the following types of resources exist:
122 %Resource | Documented in Section | Description
123 --------------- | --------------------- | -----------
124 AS | | Every Autonomous System (AS) may contain one or more AS.
125 host | \ref pf_host | This entity carries out the actual computation. For this reason, it contains processors (with potentially multiple cores).
126 router | \ref pf_router | In SimGrid, routers are used to provide helpful information to routing algorithms. Routers may also act as gateways, connecting several autonomous systems with each other.
127 link | \ref pf_link | In SimGrid, (network)links define a connection between two or potentially even more resources. Every link has a bandwidth and a latency and may potentially experience congestion.
128 cluster | \ref pf_cluster | In SimGrid, clusters were introduced to model large and homogenous environments. They are not really a resource by themselves - technically, they are only a shortcut, as they will internally set up all the hosts, network and routing for you, i.e., using this resource, one can easily setup thousands of hosts and links in a few lines of code. Each cluster is itself an AS.
130 As it is desirable to interconnect these resources, a routing has to
131 be defined. The AS is supposed to be Autonomous, hence this has to be
132 done at the AS level. The AS handles two different types of entities
133 (<b>host/router</b> and <b>AS</b>). However, the user is responsible
134 to define routes between those resources, otherwise entities will be
135 unconnected and therefore unreachable from other entities. Although
136 several routing algorithms are built into SimGrid (see \ref pf_rm),
137 you might encounter a case where you want to define routes manually
138 (for instance, due to specific requirements of your platform).
140 There are three tags to use:
141 \li <b>ASroute</b>: to define routes between two <b>AS</b>
142 \li <b>route</b>: to define routes between two <b>host/router</b>
143 \li <b>bypassRoute</b>: to define routes between two <b>AS</b> that
144 will bypass default routing (as specified by the ``routing`` attribute
145 supplied to ``<AS>``, see above).
147 Here is an illustration of these concepts:
149 
150 Circles represent processing units and squares represent network routers. Bold
151 lines represent communication links. AS2 models the core of a national
152 network interconnecting a small flat cluster (AS4) and a larger
153 hierarchical cluster (AS5), a subset of a LAN (AS6), and a set of peers
154 scattered around the world (AS7).
156 \section pf_pftags Resource description
158 \subsection pf_As Platform: The <AS> tag
160 The concept of an AS was already outlined above (Section \ref pf_basics);
161 recall that the AS is so important because it groups other resources (such
162 as routers/hosts) together (in fact, these resources must be contained by
165 Available attributes :
167 Attribute name | Mandatory | Values | Description
168 --------------- | --------- | ------ | -----------
169 id | yes | String | The identifier of an AS; facilitates referring to this AS. ID must be unique.
170 routing | yes | Full\| Floyd\| Dijkstra\| DijkstraCache\| None\| Vivaldi\| Cluster | See Section \ref pf_rm for details.
175 <AS id="AS0" routing="Full">
176 <host id="host1" speed="1000000000"/>
177 <host id="host2" speed="1000000000"/>
178 <link id="link1" bandwidth="125000000" latency="0.000100"/>
179 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
183 In this example, AS0 contains two hosts (host1 and host2). The route
184 between the hosts goes through link1.
186 \subsection pf_Cr Computing resources: hosts, clusters and peers.
188 \subsubsection pf_host <host/>
190 A <b>host</b> represents a computer/node card. Every host is able to execute
191 code and it can send and receive data to/from other hosts. Most importantly,
192 a host can contain more than 1 core.
196 Attribute name | Mandatory | Values | Description
197 --------------- | --------- | ------ | -----------
198 id | yes | String | The identifier of the host. facilitates referring to this AS.
199 speed | yes | double (must be > 0.0) | Computational power of every core of this host in FLOPS. Must be larger than 0.0.
200 core | no | int (Default: 1) | The number of cores of this host. If more than one core is specified, the "speed" parameter refers to every core, i.e., the total computational power is no_of_cores*speed.<br /> If 6 cores are specified, up to 6 tasks can be executed without sharing the computational power; if more than 6 tasks are executed, computational power will be shared among these tasks. <br /> <b>Warning:</b> Although functional, this model was never scientifically assessed.
201 availability_file| no | string | (Relative or absolute) filename to use as input; must contain availability traces for this host. The syntax of this file is defined below. <br /> <b>Note:</b> The filename must be specified with your system's format.
202 state_file | no | string | Same mechanism as availability_file.<br /> <b>Note:</b> The filename must be specified with your system's format.
203 coordinates | no | string | Must be provided when choosing the Vivaldi, coordinate-based routing model for the AS the host belongs to. More details can be found in the Section \ref pf_P2P_tags.
204 pstate | no | double (Default: 0.0) | FIXME: Not yet documented.
206 ### Possible children: ###
208 Tag name | Description | Documentation
209 ------------ | ----------- | -------------
210 \<mount/\> | Defines mounting points between some storage resource and the host. | \ref pf_storage_entity_mount
211 \<prop/\> | The prop tag allows you to define additional information on this host following the attribute/value schema. You may want to use it to give information to the tool you use for rendering your simulation, for example. | N/A
216 <host id="host1" speed="1000000000"/>
217 <host id="host2" speed="1000000000">
218 <prop id="color" value="blue"/>
219 <prop id="rendershape" value="square"/>
224 \anchor pf_host_dynamism
225 ### Expressing dynamism ###
227 SimGrid provides mechanisms to change a hosts' availability over
228 time, using the ``availability_file`` attribute to the ``\<host\>`` tag
229 and a separate text file whose syntax is exemplified below.
231 #### Adding a trace file ####
234 <platform version="4">
235 <host id="bob" speed="500Gf" availability_file="bob.trace" />
239 #### Example of "bob.trace" file ####
248 Let us begin to explain this example by looking at line 2. (Line 1 will become clear soon).
249 The first column describes points in time, in this case, time 0. The second column
250 describes the relative amount of power this host is able to deliver (relative
251 to the maximum performance specified in the ``\<host\>`` tag). (Clearly, the
252 second column needs to contain values that are not smaller than 0 and not larger than 1).
253 In this example, our host will deliver 500 Mflop/s at time 0, as 500 Mflop/s is the
254 maximum performance of this host. At time 11.0, it will
255 deliver half of its maximum performance, i.e., 250 Mflop/s until time 20.0 when it will
256 will start delivering 80\% of its power. In this example, this amounts to 400 Mflop/s.
258 Since the periodicity in line 1 was set to be 1.0, i.e., 1 timestep, this host will
259 continue to provide 500 Mflop/s from time 21. From time 32 it will provide 250 MFlop/s and so on.
261 ### Changing initial state ###
263 It is also possible to specify whether the host is up or down by setting the
264 ``state`` attribute to either <b>ON</b> (default value) or <b>OFF</b>.
266 #### Example: Expliciting the default value "ON" ####
269 <platform version="4">
270 <host id="bob" speed="500Gf" state="ON" />
274 If you want this host to be unavailable, simply substitute ON with OFF.
276 \anchor pf_host_churn
277 ### Expressing churn ###
279 To express the fact that a host can change state over time (as in P2P
280 systems, for instance), it is possible to use a file describing the time
281 at which the host is turned on or off. An example of the content
282 of such a file is presented below.
284 #### Adding a state file ####
287 <platform version="4">
288 <host id="bob" power="500Gf" state_file="bob.fail" />
292 #### Example of "bob.fail" file ####
300 A zero value means <b>down</b> (i.e., OFF) while a positive one means <b>up and
301 running</b> (i.e., ON). From time 0.0 to time 1.0, the host is on as usual. At time 1.0, it is
302 turned off and at time 2.0, it is turned on again until time 12 (2 plus the
303 periodicity 10). It will be turned off again at time 13.0 until time 23.0, and
307 \subsubsection pf_cluster <cluster>
309 ``<cluster />`` represents a machine-cluster. It is most commonly used
310 when one wants to define many hosts and a network quickly. Technically,
311 ``cluster`` is a meta-tag: <b>from the inner SimGrid point of
312 view, a cluster is an AS where some optimized routing is defined</b>.
313 The default inner organization of the cluster is as follow:
319 ____________|__________|_____________ backbone
321 l0| l1| l2| l97| l96 | | l99
327 Here, a set of <b>host</b>s is defined. Each of them has a <b>link</b>
328 to a central backbone (backbone is a link itself, as a link can
329 be used to represent a switch, see the switch / link section
330 below for more details about it). A <b>router</b> allows to connect a
331 <b>cluster</b> to the outside world. Internally,
332 SimGrid treats a cluster as an AS containing all hosts: the router is the default
333 gateway for the cluster.
335 There is an alternative organization, which is as follows:
349 The principle is the same, except that there is no backbone. This representation
350 can be obtained easily: just do not set the bb_* attributes.
353 Attribute name | Mandatory | Values | Description
354 --------------- | --------- | ------ | -----------
355 id | yes | string | The identifier of the cluster. Facilitates referring to this cluster.
356 prefix | yes | string | Each node of the cluster has to have a name. This name will be prefixed with this prefix.
357 suffix | yes | string | Each node of the cluster will be suffixed with this suffix
358 radical | yes | string | Regexp used to generate cluster nodes name. Syntax: "10-20" will give you 11 machines numbered from 10 to 20, "10-20;2" will give you 12 machines, one with the number 2, others numbered as before. The produced number is concatenated between prefix and suffix to form machine names.
359 speed | yes | int | Same as the ``speed`` attribute of the ``\<host\>`` tag.
360 core | no | int (default: 1) | Same as the ``core`` attribute of the ``\<host\>`` tag.
361 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_link "link section" for syntax/details.
362 lat | yes | int | Latency for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
363 sharing_policy | no | string | Sharing policy for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
364 bb_bw | no | int | Bandwidth for backbone (if any). See <b>link</b> section for syntax/details. If bb_bw and bb_lat (see below) attributes are omitted, no backbone is created (alternative cluster architecture <b>described before</b>).
365 bb_lat | no | int | Latency for backbone (if any). See <b>link</b> section for syntax/details. If bb_lat and bb_bw (see above) attributes are omitted, no backbone is created (alternative cluster architecture <b>described before</b>).
366 bb_sharing_policy | no | string | Sharing policy for the backbone (if any). See <b>link</b> section for syntax/details.
367 limiter_link | no | int | Bandwidth for limiter link (if any). This adds a specific link for each node, to set the maximum bandwidth reached when communicating in both directions at the same time. In theory this value should be 2*bw for fullduplex links, but in reality this might be less. This value will depend heavily on the communication model, and on the cluster's hardware, so no default value can be set, this has to be measured. More details can be obtained in <a href="https://hal.inria.fr/hal-00919507/"> "Toward Better Simulation of MPI Applications on Ethernet/TCP Networks"</a>
368 loopback_bw | no | int | Bandwidth for loopback (if any). See <b>link</b> section for syntax/details. If loopback_bw and loopback_lat (see below) attributes are omitted, no loopback link is created and all intra-node communication will use the main network link of the node. Loopback link is a \ref pf_sharing_policy_fatpipe "\b FATPIPE".
369 loopback_lat | no | int | Latency for loopback (if any). See <b>link</b> section for syntax/details. See loopback_bw for more info.
370 topology | no | FLAT\|TORUS\|FAT_TREE\|DRAGONFLY (default: FLAT) | Network topology to use. SimGrid currently supports FLAT (with or without backbone, as described before), <a href="http://en.wikipedia.org/wiki/Torus_interconnect">TORUS </a>, FAT_TREE, and DRAGONFLY attributes for this tag.
371 topo_parameters | no | string | Specific parameters to pass for the topology defined in the topology tag. For torus networks, comma-separated list of the number of nodes in each dimension of the torus. For fat trees, refer to \ref simgrid::kernel::routing::AsClusterFatTree "AsClusterFatTree documentation". For dragonfly, refer to \ref simgrid::kernel::routing::AsClusterDragonfly "AsClusterDragonfly documentation".
374 the router name is defined as the resulting String in the following
378 router_name = prefix + clusterId + _router + suffix;
382 #### Cluster example ####
384 Consider the following two (and independent) uses of the ``cluster`` tag:
387 <cluster id="my_cluster_1" prefix="" suffix="" radical="0-262144"
388 speed="1e9" bw="125e6" lat="5E-5"/>
390 <cluster id="my_cluster_2" prefix="c-" suffix=".me" radical="0-99"
391 speed="1e9" bw="125e6" lat="5E-5"
392 bb_bw="2.25e9" bb_lat="5E-4"/>
395 The second example creates one router and 100 machines with the following names:
397 c-my_cluster_2_router.me
405 \subsubsection pf_cabinet <cabinet>
408 This tag is only available when the routing mode of the AS
409 is set to ``Cluster``.
411 The ``<cabinet />`` tag is, like the \ref pf_cluster "<cluster>" tag,
412 a meta-tag. This means that it is simply a shortcut for creating a set of (homogenous) hosts and links quickly;
413 unsurprisingly, this tag was introduced to setup cabinets in data centers quickly. Unlike
414 <cluster>, however, the <cabinet> assumes that you create the backbone
415 and routers yourself; see our examples below.
419 Attribute name | Mandatory | Values | Description
420 --------------- | --------- | ------ | -----------
421 id | yes | string | The identifier of the cabinet. Facilitates referring to this cluster.
422 prefix | yes | string | Each node of the cabinet has to have a name. This name will be prefixed with this prefix.
423 suffix | yes | string | Each node of the cabinet will be suffixed with this suffix
424 radical | yes | string | Regexp used to generate cabinet nodes name. Syntax: "10-20" will give you 11 machines numbered from 10 to 20, "10-20;2" will give you 12 machines, one with the number 2, others numbered as before. The produced number is concatenated between prefix and suffix to form machine names.
425 speed | yes | int | Same as the ``speed`` attribute of the \ref pf_host "<host>" tag.
426 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_link "link section" for syntax/details.
427 lat | yes | int | Latency for the links between nodes and backbone (if any). See the \ref pf_link "link section" for syntax/details.
430 Please note that as of now, it is impossible to change attributes such as,
431 amount of cores (always set to 1), the initial state of hosts/links
432 (always set to ON), the sharing policy of the links (always set to \ref pf_sharing_policy_fullduplex "FULLDUPLEX").
436 The following example was taken from ``examples/platforms/meta_cluster.xml`` and
437 shows how to use the cabinet tag.
440 <AS id="my_cluster1" routing="Cluster">
441 <cabinet id="cabinet1" prefix="host-" suffix=".cluster1"
442 speed="1Gf" bw="125MBps" lat="100us" radical="1-10"/>
443 <cabinet id="cabinet2" prefix="host-" suffix=".cluster1"
444 speed="1Gf" bw="125MBps" lat="100us" radical="11-20"/>
445 <cabinet id="cabinet3" prefix="host-" suffix=".cluster1"
446 speed="1Gf" bw="125MBps" lat="100us" radical="21-30"/>
448 <backbone id="backbone1" bandwidth="2.25GBps" latency="500us"/>
453 Please note that you must specify the \ref pf_backbone "<backbone>"
454 tag by yourself; this is not done automatically and there are no checks
455 that ensure this backbone was defined.
457 The hosts generated in the above example are named host-1.cluster, host-2.cluster1
461 \subsubsection pf_peer The <peer> tag
463 This tag represents a peer, as in Peer-to-Peer (P2P) networks. However, internally,
464 SimGrid transforms a peer into an AS (similar to Cluster). Hence, this tag
465 is virtually only a shortcut that comes with some pre-defined resources
466 and values. These are:
468 \li A tiny AS whose routing type is cluster is created
470 \li Two links: One for download and one for upload. This is
471 convenient to use and simulate stuff under the last mile model (e.g., ADSL peers).
472 \li It connects the two links to the host
473 \li It creates a router (a gateway) that serves as an entry point for this peer zone.
474 This router has coordinates.
478 Attribute name | Mandatory | Values | Description
479 --------------- | --------- | ------ | -----------
480 id | yes | string | The identifier of the peer. Facilitates referring to this peer.
481 speed | yes | int | See the description of the ``host`` tag for this attribute
482 bw_in | yes | int | Bandwidth downstream
483 bw_out | yes | int | Bandwidth upstream
484 lat | yes | double | Latency for both up- and downstream, in seconds.
485 coordinates | no | string | Coordinates of the gateway for this peer. Example value: 12.8 14.4 6.4
486 sharing_policy | no | SHARED\|FULLDUPLEX (default: FULLDUPLEX) | Sharing policy for links. See <b>link</b> description for details.
487 availability_file| no | string | Availability file for the peer. Same as host availability file. See <b>host</b> description for details.
488 state_file | no | string | State file for the peer. Same as host state file. See <b>host</b> description for details.
490 Internally, SimGrid transforms any ``\<peer/\>`` construct such as
493 coordinates="12.8 14.4 6.4"
499 into an ``\<AS\>`` (see Sections \ref pf_basics and \ref pf_As). In fact, this example of the ``\<peer/\>`` tag
500 is completely equivalent to the following declaration:
503 <AS id="as_FOO" routing="Cluster">
504 <host id="peer_FOO" speed="1.5Gf"/>
505 <link id="link_FOO_UP" bandwidth="2.25GBps" latency="500us"/>
506 <link id="link_FOO_DOWN" bandwidth="2.25GBps" latency="500us"/>
507 <router id="router_FOO" coordinates="25.5 9.4 1.4"/>
508 <host_link id="peer_FOO" up="link_FOO_UP" down="link_FOO_DOWN"/>
513 \subsection pf_ne Network equipments: links and routers
515 There are two tags at all times available to represent network entities and
516 several other tags that are available only in certain contexts.
517 1. ``<link>``: Represents a entity that has a limited bandwidth, a
518 latency, and that can be shared according to TCP way to share this
521 The concept of links in SimGrid may not be intuitive, as links are not
522 limited to connecting (exactly) two entities; in fact, you can have more than
523 two equipments connected to it. (In graph theoretical terms: A link in
524 SimGrid is not an edge, but a hyperedge)
526 2. ``<router/>``: Represents an entity that a message can be routed
527 to, but that is unable to execute any code. In SimGrid, routers have also
528 no impact on the performance: Routers do not limit any bandwidth nor
529 do they increase latency. As a matter of fact, routers are (almost) ignored
530 by the simulator when the simulation has begun.
532 3. ``<backbone/>``: This tag is only available when the containing AS is
533 used as a cluster (i.e., mode="Cluster")
536 If you want to represent an entity like a switch, you must use ``<link>`` (see section). Routers are used
537 to run some routing algorithm and determine routes (see Section \ref pf_routing for details).
539 \subsubsection pf_router <router/>
541 As said before, <b>router</b> is used only to give some information
542 for routing algorithms. So, it does not have any attributes except :
546 Attribute name | Mandatory | Values | Description
547 --------------- | --------- | ------ | -----------
548 id | yes | string | The identifier of the router to be used when referring to it.
549 coordinates | no | string | Must be provided when choosing the Vivaldi, coordinate-based routing model for the AS the router belongs to. More details can be found in the Section \ref pf_P2P_tags.
554 <router id="gw_dc1_horizdist"/>
557 \subsubsection pf_link <link/>
559 Network links can represent one-hop network connections. They are
560 characterized by their id and their bandwidth; links can (but may not) be subject
565 Attribute name | Mandatory | Values | Description
566 --------------- | --------- | ------ | -----------
567 id | yes | string | The identifier of the link to be used when referring to it.
568 bandwidth | yes | int | Maximum bandwidth for this link, given in bytes/s
569 latency | no | double (default: 0.0) | Latency for this link.
570 sharing_policy | no | \ref sharing_policy_shared "SHARED"\|\ref pf_sharing_policy_fatpipe "FATPIPE"\|\ref pf_sharing_policy_fullduplex "FULLDUPLEX" (default: SHARED) | Sharing policy for the link.
571 state | no | ON\|OFF (default: ON) | Allows you to to turn this link on or off (working / not working)
572 bandwidth_file | no | string | Allows you to use a file as input for bandwidth.
573 latency_file | no | string | Allows you to use a file as input for latency.
574 state_file | no | string | Allows you to use a file as input for states.
577 #### Possible shortcuts for ``latency`` ####
579 When using the latency attribute, you can specify the latency by using the scientific
580 notation or by using common abbreviations. For instance, the following three tags
584 <link id="LINK1" bandwidth="125000000" latency="5E-6"/>
585 <link id="LINK1" bandwidth="125000000" latency="5us"/>
586 <link id="LINK1" bandwidth="125000000" latency="0.000005"/>
589 Here, the second tag uses "us", meaning "microseconds". Other shortcuts are:
591 Name | Abbreviation | Time (in seconds)
592 ---- | ------------ | -----------------
593 Week | w | 7 * 24 * 60 * 60
594 Day | d | 24 * 60 * 60
598 Millisecond | ms | 0.001 = 10^(-3)
599 Microsecond | us | 0.000001 = 10^(-6)
600 Nanosecond | ns | 0.000000001 = 10^(-9)
601 Picosecond | ps | 0.000000000001 = 10^(-12)
603 #### Sharing policy ####
605 \anchor sharing_policy_shared
606 By default a network link is \b SHARED, i.e., if two or more data flows go
607 through a link, the bandwidth is shared fairly among all data flows. This
608 is similar to the sharing policy TCP uses.
610 \anchor pf_sharing_policy_fatpipe
611 On the other hand, if a link is defined as a \b FATPIPE,
612 each flow going through this link will be provided with the complete bandwidth,
613 i.e., no sharing occurs and the bandwidth is only limiting each flow individually.
614 Please note that this is really on a per-flow basis, not only on a per-host basis!
615 The complete bandwidth provided by this link in this mode
616 is ``number_of_flows*bandwidth``, with at most ``bandwidth`` being available per flow.
618 Using the FATPIPE mode allows to model backbones that won't affect performance
621 \anchor pf_sharing_policy_fullduplex
622 The last mode available is \b FULLDUPLEX. This means that SimGrid will
623 automatically generate two links (one carrying the suffix _UP and the other the
624 suffix _DOWN) for each ``<link>`` tag. This models situations when the direction
625 of traffic is important.
628 Transfers from one side to the other will interact similarly as
629 TCP when ACK returning packets circulate on the other direction. More
630 discussion about it is available in the description of link_ctn description.
632 In other words: The SHARED policy defines a physical limit for the bandwidth.
633 The FATPIPE mode defines a limit for each application,
634 with no upper total limit.
637 Tip: By using the FATPIPE mode, you can model big backbones that
638 won't affect performance (except latency).
643 <link id="SWITCH" bandwidth="125000000" latency="5E-5" sharing_policy="FATPIPE" />
646 #### Expressing dynamism and failures ####
648 Similar to hosts, it is possible to declare links whose state, bandwidth
649 or latency changes over time (see Section \ref pf_host_dynamism for details).
651 In the case of network links, the ``bandwidth`` and ``latency`` attributes are
652 replaced by the ``bandwidth_file`` and ``latency_file`` attributes.
653 The following XML snippet demonstrates how to use this feature in the platform
654 file. The structure of the files "link1.bw" and "link1.lat" is shown below.
657 <link id="LINK1" state_file="link1.fail" bandwidth="80000000" latency=".0001" bandwidth_file="link1.bw" latency_file="link1.lat" />
661 Even if the syntax is the same, the semantic of bandwidth and latency
662 trace files differs from that of host availability files. For bandwidth and
663 latency, the corresponding files do not
664 express availability as a fraction of the available capacity but directly in
665 bytes per seconds for the bandwidth and in seconds for the latency. This is
666 because most tools allowing to capture traces on real platforms (such as NWS)
667 express their results this way.
669 ##### Example of "link1.bw" file #####
677 In this example, the bandwidth changes repeatedly, with all changes
678 being repeated every 12 seconds.
680 At the beginning of the the simulation, the link's bandwidth is 80,000,000
681 B/s (i.e., 80 Mb/s); this value was defined in the XML snippet above.
682 After four seconds, it drops to 40 Mb/s (line 2), and climbs
683 back to 60 Mb/s after another 4 seconds (line 3). The value does not change any
684 more until the end of the period, that is, after 12 seconds have been simulated).
685 At this point, periodicity kicks in and this behavior is repeated: Seconds
686 12-16 will experience 80 Mb/s, 16-20 40 Mb/s etc.).
688 ##### Example of "link1.lat" file #####
697 In this example, the latency varies with a period of 5 seconds.
698 In the xml snippet above, the latency is initialized to be 0.0001s (100µs). This
699 value will be kept during the first second, since the latency_file contains
700 changes to this value at second one, two and three.
701 At second one, the value will be 0.001, i.e., 1ms. One second later it will
702 be adjusted to 0.01 (or 10ms) and one second later it will be set again to 1ms. The
703 value will not change until second 5, when the periodicity defined in line 1
704 kicks in. It then loops back, starting at 100µs (the initial value) for one second.
707 #### The ``<prop/>`` tag ####
709 Similar to the ``<host>`` tag, a link may also contain the ``<prop/>`` tag; see the host
710 documentation (Section \ref pf_host) for an example.
713 \subsubsection pf_backbone <backbone/>
716 This tag is <b>only available</b> when the containing AS uses the "Cluster" routing mode!
718 Using this tag, you can designate an already existing link to be a backbone.
720 Attribute name | Mandatory | Values | Description
721 --------------- | --------- | ------ | -----------
722 id | yes | string | Name of the link that is supposed to act as a backbone.
724 \subsection pf_storage Storage
727 This is a prototype version that should evolve quickly, hence this
728 is just some doc valuable only at the time of writing.
729 This section describes the storage management under SimGrid ; nowadays
730 it's only usable with MSG. It relies basically on linux-like concepts.
731 You also may want to have a look to its corresponding section in
732 @ref msg_file ; access functions are organized as a POSIX-like
735 \subsubsection pf_sto_conc Storage - Main Concepts
737 The storage facilities implemented in SimGrid help to model (and account for)
738 storage devices, such as tapes, hard-drives, CD or DVD devices etc.
739 A typical situation is depicted in the figure below:
741 \image html ./webcruft/storage_sample_scenario.png
742 \image latex ./webcruft/storage_sample_scenario.png "storage_sample_scenario" width=\textwidth
744 In this figure, two hosts called Bob and Alice are interconnected via a network
745 and each host is physically attached to a disk; it is not only possible for each host to
746 mount the disk they are attached to directly, but they can also mount disks
747 that are in a remote location. In this example, Bob mounts Alice's disk remotely
748 and accesses the storage via the network.
750 SimGrid provides 3 different entities that can be used to model setups
751 that include storage facilities:
753 Entity name | Description
754 --------------- | -----------
755 \ref pf_storage_entity_storage_type "storage_type" | Defines a template for a particular kind of storage (such as a hard-drive) and specifies important features of the storage, such as capacity, performance (read/write), contents, ... Different models of hard-drives use different storage_types (because the difference between an SSD and an HDD does matter), as they differ in some specifications (e.g., different sizes or read/write performance).
756 \ref pf_storage_entity_storage "storage" | Defines an actual instance of a storage type (disk, RAM, ...); uses a ``storage_type`` template (see line above) so that you don't need to re-specify the same details over and over again.
757 \ref pf_storage_entity_mount "mount" | Must be wrapped by a \ref pf_host tag; declares which storage(s) this host has mounted and where (i.e., the mountpoint).
760 \anchor pf_storage_content_file
761 ### %Storage Content File ###
763 In order to assess exactly how much time is spent reading from the storage,
764 SimGrid needs to know what is stored on the storage device (identified by distinct (file-)name, like in a file system)
765 and what size this content has.
768 The content file is never changed by the simulation; it is parsed once
769 per simulation and kept in memory afterwards. When the content of the
770 storage changes, only the internal SimGrid data structures change.
772 \anchor pf_storage_content_file_structure
773 #### Structure of a %Storage Content File ####
775 Here is an excerpt from two storage content file; if you want to see the whole file, check
776 the file ``examples/platforms/content/storage_content.txt`` that comes with the
779 SimGrid essentially supports two different formats: UNIX-style filepaths should
780 follow the well known format:
783 /lib/libsimgrid.so.3.6.2 12710497
787 /bin/simgrid_update_xml 5018
788 /bin/graphicator 66986
789 /bin/simgrid-colorizer 2993
794 Windows filepaths, unsurprisingly, use the windows style:
797 \Windows\avastSS.scr 41664
798 \Windows\bfsvc.exe 75264
799 \Windows\bootstat.dat 67584
800 \Windows\CoreSingleLanguage.xml 31497
802 \Windows\dchcfg64.exe 335464
803 \Windows\dcmdev64.exe 93288
807 The different file formats come at a cost; in version 3.12 (and most likely
808 in later versions, too), copying files from windows-style storages to unix-style
809 storages (and vice versa) is not supported.
811 \anchor pf_storage_content_file_create
812 #### Generate a %Storage Content File ####
814 If you want to generate a storage content file based on your own filesystem (or at least a filesystem you have access to),
815 try running this command (works only on unix systems):
818 find . -type f -exec ls -1s --block=1 {} \; 2>/dev/null | awk '{ print $2 " " $1}' > ./content.txt
821 \subsubsection pf_storage_entities The Storage Entities
823 These are the entities that you can use in your platform files to include
824 storage in your model. See also the list of our \ref pf_storage_example_files "example files";
825 these might also help you to get started.
827 \anchor pf_storage_entity_storage_type
828 #### \<storage_type\> ####
830 Attribute name | Mandatory | Values | Description
831 --------------- | --------- | ------ | -----------
832 id | yes | string | Identifier of this storage_type; used when referring to it
833 model | yes | string | For reasons of future backwards compatibility only; specifies the name of the model for the storage that should be used
834 size | yes | string | Specifies the amount of available storage space; you can specify storage like "500GiB" or "500GB" if you want. (TODO add a link to all the available abbreviations)
835 content | yes | string | Path to a \ref pf_storage_content_file "Storage Content File" on your system. This file must exist.
836 content_type | no | ("txt_unix"\|"txt_win") | Determines which kind of filesystem you're using; make sure the filenames (stored in that file, see \ref pf_storage_content_file_structure "Storage Content File Structure"!)
838 This tag must contain some predefined model properties, specified via the <model_prop> tag. Here is a list,
839 see below for an example:
841 Property id | Mandatory | Values | Description
842 --------------- | --------- | ------ | -----------
843 Bwrite | yes | string | Bandwidth for write access; in B/s (but you can also specify e.g. "30MBps")
844 Bread | yes | string | Bandwidth for read access; in B/s (but you can also specify e.g. "30MBps")
845 Bconnexion | yes | string | Throughput (of the storage connector) in B/s.
848 A storage_type can also contain the <b><prop></b> tag. The <prop> tag allows you
849 to associate additional information to this <storage_type> and follows the
850 attribute/value schema; see the example below. You may want to use it to give information to
851 the tool you use for rendering your simulation, for example.
853 Here is a complete example for the ``storage_type`` tag:
855 <storage_type id="single_HDD" model="linear_no_lat" size="4000" content_type="txt_unix">
856 <model_prop id="Bwrite" value="30MBps" />
857 <model_prop id="Bread" value="100MBps" />
858 <model_prop id="Bconnection" value="150MBps" />
859 <prop id="Brand" value="Western Digital" />
863 \anchor pf_storage_entity_storage
864 #### <storage> ####
866 ``storage`` attributes:
868 Attribute name | Mandatory | Values | Description
869 -------------- | --------- | ------ | -----------
870 id | yes | string | Identifier of this ``storage``; used when referring to it
871 typeId | yes | string | Here you need to refer to an already existing \ref pf_storage_entity_storage_type "\<storage_type\>"; the storage entity defined by this tag will then inherit the properties defined there.
872 attach | yes | string | Name of a host (see Section \ref pf_host) to which this storage is <i>physically</i> attached to (e.g., a hard drive in a computer)
873 content | no | string | When specified, overwrites the content attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
874 content_type | no | string | When specified, overwrites the content_type attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
876 Here are two examples:
879 <storage id="Disk1" typeId="single_HDD" attach="bob" />
881 <storage id="Disk2" typeId="single_SSD"
882 content="content/win_storage_content.txt"
883 content_type="txt_windows" attach="alice" />
886 The first example is straightforward: A disk is defined and called "Disk1"; it is
887 of type "single_HDD" (shown as an example of \ref pf_storage_entity_storage_type "\<storage_type\>" above) and attached
888 to a host called "bob" (the definition of this host is omitted here).
890 The second storage is called "Disk2", is still of the same type as Disk1 but
891 now specifies a new content file (so the contents will be different from Disk1)
892 and the filesystem uses the windows style; finally, it is attached to a second host,
893 called alice (which is again not defined here).
895 \anchor pf_storage_entity_mount
896 #### <mount> ####
899 | Attribute name | Mandatory | Values | Description |
900 | ---------------- | ----------- | -------- | ------------- |
901 | id | yes | string | Refers to a \ref pf_storage_entity_storage "<storage>" entity that will be mounted on that computer |
902 | name | yes | string | Path/location to/of the logical reference (mount point) of this disk
904 This tag must be enclosed by a \ref pf_host tag. It then specifies where the mountpoint of a given storage device (defined by the ``id`` attribute)
905 is; this location is specified by the ``name`` attribute.
907 Here is a simple example, taken from the file ``examples/platform/storage.xml``:
910 <storage_type id="single_SSD" model="linear_no_lat" size="500GiB">
911 <model_prop id="Bwrite" value="60MBps" />
912 <model_prop id="Bread" value="200MBps" />
913 <model_prop id="Bconnection" value="220MBps" />
916 <storage id="Disk2" typeId="single_SSD"
917 content="content/win_storage_content.txt"
918 content_type="txt_windows" attach="alice" />
919 <storage id="Disk4" typeId="single_SSD"
920 content="content/small_content.txt"
921 content_type="txt_unix" attach="denise"/>
923 <host id="alice" speed="1Gf">
924 <mount storageId="Disk2" name="c:"/>
927 <host id="denise" speed="1Gf">
928 <mount storageId="Disk2" name="c:"/>
929 <mount storageId="Disk4" name="/home"/>
933 This example is quite interesting, as the same device, called "Disk2", is mounted by
934 two hosts at the same time! Note, however, that the host called ``alice`` is actually
935 attached to this storage, as can be seen in the \ref pf_storage_entity_storage "<storage>"
936 tag. This means that ``denise`` must access this storage through the network, but SimGrid automatically takes
937 care of that for you.
939 Furthermore, this example shows that ``denise`` has mounted two storages with different
940 filesystem types (unix and windows). In general, a host can mount as many storage devices as
944 Again, the difference between ``attach`` and ``mount`` is simply that
945 an attached storage is always physically inside (or connected to) that machine;
946 for instance, a USB stick is attached to one and only one machine (where it's plugged-in)
947 but it can only be mounted on others, as mounted storage can also be a remote location.
949 ###### Example files #####
951 \verbinclude example_filelist_xmltag_mount
953 \subsubsection pf_storage_example_files Example files
955 Several examples were already discussed above; if you're interested in full examples,
956 check the the following platforms:
958 1. ``examples/platforms/storage.xml``
959 2. ``examples/platforms/remote_io.xml``
961 If you're looking for some examplary C code, you may find the source code
962 available in the directory ``examples/msg/io/`` useful.
964 \subsubsection pf_storage_examples_modelling Modelling different situations
966 The storage functionality of SimGrid is type-agnostic, that is, the implementation
967 does not presume any type of storage, such as HDDs/SSDs, RAM,
968 CD/DVD devices, USB sticks etc.
970 This allows the user to apply the simulator for a wide variety of scenarios; one
971 common scenario would be the access of remote RAM.
973 #### Modelling the access of remote RAM ####
975 How can this be achieved in SimGrid? Let's assume we have a setup where three hosts
976 (HostA, HostB, HostC) need to access remote RAM:
986 An easy way to model this scenario is to setup and define the RAM via the
987 \ref pf_storage_entity_storage "storage" and \ref pf_storage_entity_storage_type "storage type"
988 entities and attach it to a remote dummy host; then, every host can have their own links
989 to this host (modelling for instance certain scenarios, such as PCIe ...)
994 RAM - Dummy -- Host B
999 Now, if read from this storage, the host that mounts this storage
1000 communicates to the dummy host which reads from RAM and
1001 sends the information back.
1004 \section pf_routing Routing
1006 To achieve high performance, the routing tables used within SimGrid are
1007 static. This means that routing between two nodes is calculated once
1008 and will not change during execution. The SimGrid team chose to use this
1009 approach as it is rare to have a real deficiency of a resource;
1010 most of the time, a communication fails because the links experience too much
1011 congestion and hence, your connection stops before the timeout or
1012 because the computer designated to be the destination of that message
1015 We also chose to use shortest paths algorithms in order to emulate
1016 routing. Doing so is consistent with the reality: [RIP](https://en.wikipedia.org/wiki/Routing_Information_Protocol),
1017 [OSPF](https://en.wikipedia.org/wiki/Open_Shortest_Path_First), [BGP](https://en.wikipedia.org/wiki/Border_Gateway_Protocol)
1018 are all calculating shortest paths. They do require some time to converge, but
1019 eventually, when the routing tables have stabilized, your packets will follow
1022 \subsection pf_rm Routing models
1024 For each AS, you must define explicitly which routing model will
1025 be used. There are 3 different categories for routing models:
1027 1. \ref pf_routing_model_shortest_path "Shortest-path" based models: SimGrid calculates shortest
1028 paths and manages them. Behaves more or less like most real life
1030 2. \ref pf_routing_model_manual "Manually-entered" route models: you have to define all routes
1031 manually in the platform description file; this can become
1032 tedious very quickly, as it is very verbose.
1033 Consistent with some manually managed real life routing.
1034 3. \ref pf_routing_model_simple "Simple/fast models": those models offer fast, low memory routing
1035 algorithms. You should consider to use this type of model if
1036 you can make some assumptions about your AS.
1037 Routing in this case is more or less ignored.
1039 \subsubsection pf_raf The router affair
1041 Using routers becomes mandatory when using shortest-path based
1042 models or when using the bindings to the ns-3 packet-level
1043 simulator instead of the native analytical network model implemented
1046 For graph-based shortest path algorithms, routers are mandatory, because these
1047 algorithms require a graph as input and so we need to have source and
1048 destination for each edge.
1050 Routers are naturally an important concept ns-3 since the
1051 way routers run the packet routing algorithms is actually simulated.
1052 SimGrid's analytical models however simply aggregate the routing time
1053 with the transfer time.
1055 So why did we incorporate routers in SimGrid? Rebuilding a graph representation
1056 only from the route information turns out to be a very difficult task, because
1057 of the missing information about how routes intersect. That is why we
1058 introduced routers, which are simply used to express these intersection points.
1059 It is important to understand that routers are only used to provide topological
1062 To express this topological information, a <b>route</b> has to be
1063 defined in order to declare which link is connected to a router.
1066 \subsubsection pf_routing_model_shortest_path Shortest-path based models
1068 The following table shows all the models that compute routes using
1069 shortest-paths algorithms are currently available in SimGrid. More detail on how
1070 to choose the best routing model is given in the Section called \"\ref pf_routing_howto_choose_wisely\".
1072 | Name | Description |
1073 | --------------------------------------------------- | -------------------------------------------------------------------------- |
1074 | \ref pf_routing_model_floyd "Floyd" | Floyd routing data. Pre-calculates all routes once |
1075 | \ref pf_routing_model_dijkstra "Dijkstra" | Dijkstra routing data. Calculates routes only when needed |
1076 | \ref pf_routing_model_dijkstracache "DijkstraCache" | Dijkstra routing data. Handles some cache for already calculated routes. |
1078 All those shortest-path models are instanciated in the same way and are
1079 completely interchangeable. Here are some examples:
1081 \anchor pf_routing_model_floyd
1086 <AS id="AS0" routing="Floyd">
1088 <cluster id="my_cluster_1" prefix="c-" suffix=""
1089 radical="0-1" speed="1000000000" bw="125000000" lat="5E-5"
1090 router_id="router1"/>
1092 <AS id="AS1" routing="None">
1093 <host id="host1" speed="1000000000"/>
1096 <link id="link1" bandwidth="100000" latency="0.01"/>
1098 <ASroute src="my_cluster_1" dst="AS1"
1101 <link_ctn id="link1"/>
1107 ASroute given at the end gives a topological information: link1 is
1108 between router1 and host1.
1110 #### Example platform files ####
1112 This is an automatically generated list of example files that use the Floyd
1113 routing model (the path is given relative to SimGrid's source directory)
1115 \verbinclude example_filelist_routing_floyd
1117 \anchor pf_routing_model_dijkstra
1120 #### Example platform files ####
1122 This is an automatically generated list of example files that use the Dijkstra
1123 routing model (the path is given relative to SimGrid's source directory)
1125 \verbinclude example_filelist_routing_dijkstra
1129 <AS id="AS_2" routing="Dijkstra">
1130 <host id="AS_2_host1" speed="1000000000"/>
1131 <host id="AS_2_host2" speed="1000000000"/>
1132 <host id="AS_2_host3" speed="1000000000"/>
1133 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1134 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1135 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1136 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1137 <router id="central_router"/>
1138 <router id="AS_2_gateway"/>
1139 <!-- routes providing topological information -->
1140 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1141 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1142 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1143 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1147 \anchor pf_routing_model_dijkstracache
1148 ### DijkstraCache ###
1150 DijkstraCache example:
1152 <AS id="AS_2" routing="DijkstraCache">
1153 <host id="AS_2_host1" speed="1000000000"/>
1155 (platform unchanged compared to upper example)
1158 #### Example platform files ####
1160 This is an automatically generated list of example files that use the DijkstraCache
1161 routing model (the path is given relative to SimGrid's source directory):
1163 Editor's note: At the time of writing, no platform file used this routing model - so
1164 if there are no example files listed here, this is likely to be correct.
1166 \verbinclude example_filelist_routing_dijkstra_cache
1168 \subsubsection pf_routing_model_manual Manually-entered route models
1170 | Name | Description |
1171 | ---------------------------------- | ------------------------------------------------------------------------------ |
1172 | \ref pf_routing_model_full "Full" | You have to enter all necessary routers manually; that is, every single route. This may consume a lot of memory when the XML is parsed and might be tedious to write; i.e., this is only recommended (if at all) for small platforms. |
1174 \anchor pf_routing_model_full
1179 <AS id="AS0" routing="Full">
1180 <host id="host1" speed="1000000000"/>
1181 <host id="host2" speed="1000000000"/>
1182 <link id="link1" bandwidth="125000000" latency="0.000100"/>
1183 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
1187 #### Example platform files ####
1189 This is an automatically generated list of example files that use the Full
1190 routing model (the path is given relative to SimGrid's source directory):
1192 \verbinclude example_filelist_routing_full
1194 \subsubsection pf_routing_model_simple Simple/fast models
1196 | Name | Description |
1197 | ---------------------------------------- | ------------------------------------------------------------------------------ |
1198 | \ref pf_routing_model_cluster "Cluster" | This is specific to the \ref pf_cluster "<cluster/>" tag and should not be used by the user, as several assumptions are made. |
1199 | \ref pf_routing_model_none "None" | No routing at all. Unless you know what you're doing, avoid using this mode in combination with a non-constant network model. |
1200 | \ref pf_routing_model_vivaldi "Vivaldi" | Perfect when you want to use coordinates. Also see the corresponding \ref pf_P2P_tags "P2P section" below. |
1202 \anchor pf_routing_model_cluster
1206 In this mode, the \ref pf_cabinet "<cabinet/>" tag is available.
1208 #### Example platform files ####
1210 This is an automatically generated list of example files that use the Cluster
1211 routing model (the path is given relative to SimGrid's source directory):
1213 \verbinclude example_filelist_routing_cluster
1215 \anchor pf_routing_model_none
1218 This model does exactly what it's name advertises: Nothing. There is no routing
1219 available within this model and if you try to communicate within the AS that
1220 uses this model, SimGrid will fail unless you have explicitly activated the
1221 \ref options_model_select_network_constant "Constant Network Model" (this model charges
1222 the same for every single communication). It should
1223 be noted, however, that you can still attach an \ref pf_routing_tag_asroute "ASroute",
1224 as is demonstrated in the example below:
1226 \verbinclude platforms/cluster_and_one_host.xml
1228 #### Example platform files ####
1230 This is an automatically generated list of example files that use the None
1231 routing model (the path is given relative to SimGrid's source directory):
1233 \verbinclude example_filelist_routing_none
1236 \anchor pf_routing_model_vivaldi
1239 For more information on how to use the [Vivaldi Coordinates](https://en.wikipedia.org/wiki/Vivaldi_coordinates),
1240 see also Section \ref pf_P2P_tags "P2P tags".
1242 For documentation on how to activate this model (as some initialization must be done
1243 in the simulator), see Section \ref options_model_network_coord "Activating Coordinate Based Routing".
1245 Note that it is possible to combine the Vivaldi routing model with other routing models;
1246 an example can be found in the file \c examples/platforms/cloud.xml. This
1247 examples models an AS using Vivaldi that contains other ASes that use different
1250 #### Example platform files ####
1252 This is an automatically generated list of example files that use the None
1253 routing model (the path is given relative to SimGrid's source directory):
1255 \verbinclude example_filelist_routing_vivaldi
1258 \subsection ps_dec Defining routes
1260 There are currently four different ways to define routes:
1262 | Name | Description |
1263 | ------------------------------------------------- | ----------------------------------------------------------------------------------- |
1264 | \ref pf_routing_tag_route "route" | Used to define route between host/router |
1265 | \ref pf_routing_tag_asroute "ASroute" | Used to define route between different AS |
1266 | \ref pf_routing_tag_bypassroute "bypassRoute" | Used to supersede normal routes as calculated by the network model between host/router; e.g., can be used to use a route that is not the shortest path for any of the shortest-path routing models. |
1267 | \ref pf_routing_tag_bypassasroute "bypassASroute" | Used in the same way as bypassRoute, but for AS |
1269 Basically all those tags will contain an (ordered) list of references
1270 to link that compose the route you want to define.
1272 Consider the example below:
1275 <route src="Alice" dst="Bob">
1276 <link_ctn id="link1"/>
1277 <link_ctn id="link2"/>
1278 <link_ctn id="link3"/>
1282 The route here from host Alice to Bob will be first link1, then link2,
1283 and finally link3. What about the reverse route? \ref pf_routing_tag_route "Route" and
1284 \ref pf_routing_tag_asroute "ASroute" have an optional attribute \c symmetrical, that can
1285 be either \c YES or \c NO. \c YES means that the reverse route is the same
1286 route in the inverse order, and is set to \c YES by default. Note that
1287 this is not the case for bypass*Route, as it is more probable that you
1288 want to bypass only one default route.
1290 For an \ref pf_routing_tag_asroute "ASroute", things are just slightly more complicated, as you have
1291 to give the id of the gateway which is inside the AS you want to access ...
1292 So it looks like this:
1295 <ASroute src="AS1" dst="AS2"
1296 gw_src="router1" gw_dst="router2">
1297 <link_ctn id="link1"/>
1301 gw == gateway, so when any message are trying to go from AS1 to AS2,
1302 it means that it must pass through router1 to get out of the AS, then
1303 pass through link1, and get into AS2 by being received by router2.
1304 router1 must belong to AS1 and router2 must belong to AS2.
1306 \subsubsection pf_linkctn <link_ctn/>
1308 This entity has only one purpose: Refer to an already existing
1309 \ref pf_link "<link/>" when defining a route, i.e., it
1310 can only occur as a child of \ref pf_routing_tag_route "<route/>"
1312 | Attribute name | Mandatory | Values | Description |
1313 | --------------- | --------- | ------ | ----------- |
1314 | id | yes | String | The identifier of the link that should be added to the route. |
1315 | direction | maybe | UP\|DOWN | If the link referenced by \c id has been declared as \ref pf_sharing_policy_fullduplex "FULLDUPLEX", this indicates which direction the route traverses through this link: UP or DOWN. If you don't use FULLDUPLEX, do not use this attribute or SimGrid will not find the right link.
1317 #### Example Files ####
1319 This is an automatically generated list of example files that use the \c <link_ctn/>
1320 entity (the path is given relative to SimGrid's source directory):
1322 \verbinclude example_filelist_xmltag_linkctn
1324 \subsubsection pf_routing_tag_asroute ASroute
1326 The purpose of this entity is to define a route between two ASes.
1327 This is mainly useful when you're in the \ref pf_routing_model_full "Full routing model".
1329 #### Attributes ####
1331 | Attribute name | Mandatory | Values | Description |
1332 | --------------- | --------- | ------ | ----------- |
1333 | src | yes | String | The identifier of the source AS |
1334 | dst | yes | String | See the \c src attribute |
1335 | gw_src | yes | String | The gateway that will be used within the src AS; this can be any \ref pf_host "Host" or \ref pf_router "Router" defined within the src AS. |
1336 | gw_dst | yes | String | Same as \c gw_src, but with the dst AS instead. |
1337 | symmetrical | no | YES\|NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1342 <AS id="AS0" routing="Full">
1343 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
1344 radical="0-149" speed="1000000000" bw="125000000" lat="5E-5"
1345 bb_bw="2250000000" bb_lat="5E-4"/>
1347 <cluster id="my_cluster_2" prefix="c-" suffix=".me"
1348 radical="150-299" speed="1000000000" bw="125000000" lat="5E-5"
1349 bb_bw="2250000000" bb_lat="5E-4"/>
1351 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1353 <ASroute src="my_cluster_1" dst="my_cluster_2"
1354 gw_src="c-my_cluster_1_router.me"
1355 gw_dst="c-my_cluster_2_router.me">
1356 <link_ctn id="backbone"/>
1358 <ASroute src="my_cluster_2" dst="my_cluster_1"
1359 gw_src="c-my_cluster_2_router.me"
1360 gw_dst="c-my_cluster_1_router.me">
1361 <link_ctn id="backbone"/>
1366 \subsubsection pf_routing_tag_route route
1368 The principle is the same as for
1369 \ref pf_routing_tag_asroute "ASroute": The route contains a list of links that
1370 provide a path from \c src to \c dst. Here, \c src and \c dst can both be either a
1371 \ref pf_host "host" or \ref pf_router "router". This is mostly useful for the
1372 \ref pf_routing_model_full "Full routing model" as well as for the
1373 \ref pf_routing_model_shortest_path "shortest-paths" based models (as they require
1374 topological information).
1377 | Attribute name | Mandatory | Values | Description |
1378 | --------------- | --------- | ---------------------- | ----------- |
1379 | src | yes | String | The value given to the source's "id" attribute |
1380 | dst | yes | String | The value given to the destination's "id" attribute. |
1381 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1386 A route in the \ref pf_routing_model_full "Full routing model" could look like this:
1388 <route src="Tremblay" dst="Bourassa">
1389 <link_ctn id="4"/><link_ctn id="3"/><link_ctn id="2"/><link_ctn id="0"/><link_ctn id="1"/><link_ctn id="6"/><link_ctn id="7"/>
1393 A route in the \ref pf_routing_model_shortest_path "Shortest-Path routing model" could look like this:
1395 <route src="Tremblay" dst="Bourassa">
1400 You must only have one link in your routes when you're using them to provide
1401 topological information, as the routes here are simply the edges of the
1402 (network-)graph and the employed algorithms need to know which edge connects
1403 which pair of entities.
1405 \subsubsection pf_routing_tag_bypassasroute bypassASroute
1407 As said before, once you choose
1408 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1409 define some of your routes, which will be specific. You may also want
1410 to bypass some routes defined in lower level AS at an upper stage:
1411 <b>bypassASroute</b> is the tag you're looking for. It allows to
1412 bypass routes defined between already defined between AS (if you want
1413 to bypass route for a specific host, you should just use byPassRoute).
1414 The principle is the same as ASroute : <b>bypassASroute</b> contains
1415 list of links that are in the path between src and dst.
1417 #### Attributes ####
1419 | Attribute name | Mandatory | Values | Description |
1420 | --------------- | --------- | ---------------------- | ----------- |
1421 | src | yes | String | The value given to the source AS's "id" attribute |
1422 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1423 | gw_src | yes | String | The value given to the source gateway's "id" attribute; this can be any host or router within the src AS |
1424 | gw_dst | yes | String | The value given to the destination gateway's "id" attribute; this can be any host or router within the dst AS|
1425 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1430 <bypassASRoute src="my_cluster_1" dst="my_cluster_2"
1431 gw_src="my_cluster_1_router"
1432 gw_dst="my_cluster_2_router">
1433 <link_ctn id="link_tmp"/>
1437 This example shows that link \c link_tmp (definition not displayed here) directly
1438 connects the router \c my_cluster_1_router in the source cluster to the router
1439 \c my_cluster_2_router in the destination router. Additionally, as the \c symmetrical
1440 attribute was not given, this route is presumed to be symmetrical.
1442 \subsubsection pf_routing_tag_bypassroute bypassRoute
1444 As said before, once you choose
1445 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1446 define some of your routes, which will be specific. You may also want
1447 to bypass some routes defined in lower level AS at an upper stage :
1448 <b>bypassRoute</b> is the tag you're looking for. It allows to bypass
1449 routes defined between <b>host/router</b>. The principle is the same
1450 as route : <b>bypassRoute</b> contains list of links references of
1451 links that are in the path between src and dst.
1453 #### Attributes ####
1455 | Attribute name | Mandatory | Values | Description |
1456 | --------------- | --------- | ---------------------- | ----------- |
1457 | src | yes | String | The value given to the source AS's "id" attribute |
1458 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1459 | symmetrical | no | YES \| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1464 <bypassRoute src="host_1" dst="host_2">
1465 <link_ctn id="link_tmp"/>
1469 This example shows that link \c link_tmp (definition not displayed here) directly
1470 connects host \c host_1 to host \c host_2. Additionally, as the \c symmetrical
1471 attribute was not given, this route is presumed to be symmetrical.
1473 \subsection pb_baroex Basic Routing Example
1475 Let's say you have an AS named AS_Big that contains two other AS, AS_1
1476 and AS_2. If you want to make a host (h1) from AS_1 with another one
1477 (h2) from AS_2 then you'll have to proceed as follows:
1478 \li First, you have to ensure that a route is defined from h1 to the
1479 AS_1's exit gateway and from h2 to AS_2's exit gateway.
1480 \li Then, you'll have to define a route between AS_1 to AS_2. As those
1481 AS are both resources belonging to AS_Big, then it has to be done
1482 at AS_big level. To define such a route, you have to give the
1483 source AS (AS_1), the destination AS (AS_2), and their respective
1484 gateway (as the route is effectively defined between those two
1485 entry/exit points). Elements of this route can only be elements
1486 belonging to AS_Big, so links and routers in this route should be
1487 defined inside AS_Big. If you choose some shortest-path model,
1488 this route will be computed automatically.
1490 As said before, there are mainly 2 tags for routing :
1491 \li <b>ASroute</b>: to define routes between two <b>AS</b>
1492 \li <b>route</b>: to define routes between two <b>host/router</b>
1494 As we are dealing with routes between AS, it means that those we'll
1495 have some definition at AS_Big level. Let consider AS_1 contains 1
1496 host, 1 link and one router and AS_2 3 hosts, 4 links and one router.
1497 There will be a central router, and a cross-like topology. At the end
1498 of the crosses arms, you'll find the 3 hosts and the router that will
1499 act as a gateway. We have to define routes inside those two AS. Let
1500 say that AS_1 contains full routes, and AS_2 contains some Floyd
1501 routing (as we don't want to bother with defining all routes). As
1502 we're using some shortest path algorithms to route into AS_2, we'll
1503 then have to define some <b>route</b> to gives some topological
1504 information to SimGrid. Here is a file doing it all :
1507 <AS id="AS_Big" routing="Dijkstra">
1508 <AS id="AS_1" routing="Full">
1509 <host id="AS_1_host1" speed="1000000000"/>
1510 <link id="AS_1_link" bandwidth="1250000000" latency="5E-4"/>
1511 <router id="AS_1_gateway"/>
1512 <route src="AS_1_host1" dst="AS_1_gateway">
1513 <link_ctn id="AS_1_link"/>
1516 <AS id="AS_2" routing="Floyd">
1517 <host id="AS_2_host1" speed="1000000000"/>
1518 <host id="AS_2_host2" speed="1000000000"/>
1519 <host id="AS_2_host3" speed="1000000000"/>
1520 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1521 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1522 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1523 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1524 <router id="central_router"/>
1525 <router id="AS_2_gateway"/>
1526 <!-- routes providing topological information -->
1527 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1528 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1529 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1530 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1532 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1534 <ASroute src="AS_1" dst="AS_2"
1535 gw_src="AS_1_gateway"
1536 gw_dst="AS_2_gateway">
1537 <link_ctn id="backbone"/>
1542 \section pf_other_tags Tags not (directly) describing the platform
1544 The following tags can be used inside a \<platform\> tag even if they are not
1545 directly describing the platform:
1546 \li \ref pf_config "config": it allows you to pass some configuration stuff like, for
1547 example, the network model and so on. It follows the
1548 \li \ref pf_include "include": allows you to include another file into the current one.
1550 \subsection pf_config config
1552 The only purpose of this tag is to contain the \c prop tags, as described below.
1553 These tags will then configure the options as described by Section \ref options.
1556 #### Attributes ####
1558 | Attribute name | Mandatory | Values | Description |
1559 | --------------- | --------- | ---------------------- | ----------- |
1560 | id | yes | String | The identifier of the config tag when referring to id; this is basically useless, though. |
1562 #### Possible children ####
1564 Tag name | Description | Documentation
1565 ------------ | ----------- | -------------
1566 \<prop/\> | The prop tag allows you to define different configuration options following the attribute/value schema. See the \ref options page. | N/A
1571 <?xml version='1.0'?>
1572 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1573 <platform version="4">
1574 <config id="General">
1575 <prop id="maxmin/precision" value="0.000010"></prop>
1576 <prop id="cpu/optim" value="TI"></prop>
1577 <prop id="host/model" value="compound"></prop>
1578 <prop id="network/model" value="SMPI"></prop>
1579 <prop id="path" value="~/"></prop>
1580 <prop id="smpi/bw-factor" value="65472:0.940694;15424:0.697866;9376:0.58729"></prop>
1583 <AS id="AS0" routing="Full">
1587 \subsection pf_include include
1589 Even if it can be used in other contexts, this tag was originally created
1590 to be used with \ref pf_trace. The idea was to have a file describing the
1591 platform, and another file attaching traces of a given period to the platform.
1593 The drawback is that the file chuncks that will be included do not
1594 constitute valid XML files. This may explain why this feature was never really
1595 used in practice (as far as we know). Other mechanisms, such as the ability to load
1596 several platform files one after the other, could be considered in the future.
1598 In the meanwhile, the \c include tag allows you to import other platforms into your
1599 local file. This is done with the intention to help people
1600 combine their different AS and provide new platforms. Those files
1601 should contain XML that consists of
1602 \ref pf_include "include", \ref pf_cluster "cluster", \ref pf_peer "peer", \ref pf_As "AS", \ref pf_trace "trace", \ref pf_trace "tags".
1604 Do not forget to close the tag to make it work, or you will end up with an invalid XML file.
1606 #### Attributes ####
1608 | Attribute name | Mandatory | Values | Description |
1609 | --------------- | --------- | ---------------------- | ----------- |
1610 | file | yes | String | Filename of the path you want to include with either relative or absolute path. |
1615 The following example includes two files, clusterA.xml and clusterB.xml and
1616 combines them two one platform file; all hosts, routers etc. defined in
1617 each of them will then be usable.
1620 <?xml version='1.0'?>
1621 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid/simgrid.dtd">
1622 <platform version="4">
1623 <AS id="main" routing="Full">
1624 <include file="clusterA.xml"></include>
1625 <include file="clusterB.xml"></include>
1630 \subsection pf_trace trace and trace_connect
1632 Both tags are an alternate way to pass files containing information on
1633 availability, state etc. to an entity. (See also, for instance, Section \ref
1634 pf_host_churn "Churn", as described for the host entity.) Instead of referring
1635 to the file directly in the host, link, or cluster tag, you proceed by defining
1636 a trace with an id corresponding to a file, later a host/link/cluster, and
1637 finally using trace_connect you say that the file trace must be used by the
1644 <AS id="AS0" routing="Full">
1645 <host id="bob" speed="1000000000"/>
1647 <trace id="myTrace" file="bob.trace" periodicity="1.0"/>
1648 <trace_connect trace="myTrace" element="bob" kind="POWER"/>
1652 The order here is important. \c trace_connect must come
1653 after the elements \c trace and \c host, as both the host
1654 and the trace definition must be known when \c trace_connect
1655 is parsed; the order of \c trace and \c host is arbitrary.
1658 #### \c trace attributes ####
1661 | Attribute name | Mandatory | Values | Description |
1662 | --------------- | --------- | ---------------------- | ----------- |
1663 | id | yes | String | Identifier of this trace; this is the name you pass on to \c trace_connect. |
1664 | file | no | String | Filename of the file that contains the information - the path must follow the style of your OS. You can omit this, but then you must specifiy the values inside of <trace> and </trace> - see the example below. |
1665 | trace_periodicity | yes | String | This is the same as for \ref pf_host "hosts" (see there for details) |
1667 Here is an example of trace when no file name is provided:
1670 <trace id="myTrace" periodicity="1.0">
1677 #### \c trace_connect attributes ####
1679 | Attribute name | Mandatory | Values | Description |
1680 | --------------- | --------- | ---------------------- | ----------- |
1681 | kind | no | HOST_AVAIL\|POWER\|<br/>LINK_AVAIL\|BANDWIDTH\|LATENCY (Default: HOST_AVAIL) | Describes the kind of trace. |
1682 | trace | yes | String | Identifier of the referenced trace (specified of the trace's \c id attribute) |
1683 | element | yes | String | The identifier of the referenced entity as given by its \c id attribute |
1685 \section pf_hints Hints, tips and frequently requested features
1687 Now you should know at least the syntax and be able to create a
1688 platform by your own. However, after having ourselves wrote some platforms, there
1689 are some best practices you should pay attention to in order to
1690 produce good platform and some choices you can make in order to have
1691 faster simulations. Here's some hints and tips, then.
1693 @subsection Finding the platform example that you need
1695 Most platform files that we ship are in the @c examples/platforms
1696 folder. The good old @c grep tool can find the examples you need when
1697 wondering on a specific XML tag. Here is an example session searching
1698 for @ref pf_trace "trace_connect":
1701 % cd examples/platforms
1702 % grep -R -i -n --include="*.xml" "trace_connect" .
1703 ./two_hosts_platform_with_availability_included.xml:26:<trace_connect kind="SPEED" trace="A" element="Cpu A"/>
1704 ./two_hosts_platform_with_availability_included.xml:27:<trace_connect kind="HOST_AVAIL" trace="A_failure" element="Cpu A"/>
1705 ./two_hosts_platform_with_availability_included.xml:28:<trace_connect kind="SPEED" trace="B" element="Cpu B"/>
1706 ./two_hosts.xml:17: <trace_connect trace="Tremblay_power" element="Tremblay" kind="SPEED"/>
1709 \subsection pf_as_h AS Hierarchy
1710 The AS design allows SimGrid to go fast, because computing route is
1711 done only for the set of resources defined in this AS. If you're using
1712 only a big AS containing all resource with no AS into it and you're
1713 using Full model, then ... you'll loose all interest into it. On the
1714 other hand, designing a binary tree of AS with, at the lower level,
1715 only one host, then you'll also loose all the good AS hierarchy can
1716 give you. Remind you should always be "reasonable" in your platform
1717 definition when choosing the hierarchy. A good choice if you try to
1718 describe a real life platform is to follow the AS described in
1719 reality, since this kind of trade-off works well for real life
1722 \subsection pf_exit_as Exit AS: why and how
1723 Users that have looked at some of our platforms may have notice a
1724 non-intuitive schema ... Something like that :
1728 <AS id="AS_4" routing="Full">
1729 <AS id="exitAS_4" routing="Full">
1730 <router id="router_4"/>
1732 <cluster id="cl_4_1" prefix="c_4_1-" suffix="" radical="1-20" speed="1000000000" bw="125000000" lat="5E-5" bb_bw="2250000000" bb_lat="5E-4"/>
1733 <cluster id="cl_4_2" prefix="c_4_2-" suffix="" radical="1-20" speed="1000000000" bw="125000000" lat="5E-5" bb_bw="2250000000" bb_lat="5E-4"/>
1734 <link id="4_1" bandwidth="2250000000" latency="5E-5"/>
1735 <link id="4_2" bandwidth="2250000000" latency="5E-5"/>
1736 <link id="bb_4" bandwidth="2250000000" latency="5E-4"/>
1737 <ASroute src="cl_4_1"
1739 gw_src="c_4_1-cl_4_1_router"
1740 gw_dst="c_4_2-cl_4_2_router">
1741 <link_ctn id="4_1"/>
1742 <link_ctn id="bb_4"/>
1743 <link_ctn id="4_2"/>
1745 <ASroute src="cl_4_1"
1747 gw_src="c_4_1-cl_4_1_router"
1749 <link_ctn id="4_1"/>
1750 <link_ctn id="bb_4"/>
1752 <ASroute src="cl_4_2"
1754 gw_src="c_4_2-cl_4_2_router"
1756 <link_ctn id="4_2"/>
1757 <link_ctn id="bb_4"/>
1762 In the AS_4, you have an exitAS_4 defined, containing only one router,
1763 and routes defined to that AS from all other AS (as cluster is only a
1764 shortcut for an AS, see cluster description for details). If there was
1765 an upper AS, it would define routes to and from AS_4 with the gateway
1766 router_4. It's just because, as we did not allowed (for performances
1767 issues) to have routes from an AS to a single host/router, you have to
1768 enclose your gateway, when you have AS included in your AS, within an
1769 AS to define routes to it.
1771 \subsection pf_P2P_tags P2P or how to use coordinates
1772 SimGrid allows you to use some coordinated-based system, like vivaldi,
1773 to describe a platform. The main concept is that you have some peers
1774 that are located somewhere: this is the function of the
1775 <b>coordinates</b> of the \<peer\> or \<host\> tag. There's nothing
1776 complicated in using it, here is an example:
1779 <?xml version='1.0'?>
1780 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1781 <platform version="4">
1783 <AS id="AS0" routing="Vivaldi">
1784 <host id="100030591" coordinates="25.5 9.4 1.4" speed="1.5Gf" />
1785 <host id="100036570" coordinates="-12.7 -9.9 2.1" speed="7.3Gf" />
1787 <host id="100429957" coordinates="17.5 6.7 18.8" speed="8.3Gf" />
1792 Coordinates are then used to calculate latency (in microseconds)
1793 between two hosts by calculating the distance between the two hosts
1794 coordinates with the following formula: distance( (x1, y1, z1), (x2,
1795 y2, z2) ) = euclidian( (x1,y1), (x2,y2) ) + abs(z1) + abs(z2)
1797 In other words, we take the euclidian distance on the two first
1798 dimensions, and then add the absolute values found on the third
1799 dimension. This may seem strange, but it was found to allow better
1800 approximations of the latency matrices (see the paper describing
1803 Note that the previous example defines a routing directly between hosts but it could be also used to define a routing between AS.
1804 That is for example what is commonly done when using peers (see Section \ref pf_peer).
1806 <?xml version='1.0'?>
1807 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1808 <platform version="4">
1810 <AS id="AS0" routing="Vivaldi">
1811 <peer id="peer-0" coordinates="173.0 96.8 0.1" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us"/>
1812 <peer id="peer-1" coordinates="247.0 57.3 0.6" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1813 <peer id="peer-2" coordinates="243.4 58.8 1.4" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1817 In such a case though, we connect the AS created by the <b>peer</b> tag with the Vivaldi routing mechanism.
1818 This means that to route between AS1 and AS2, it will use the coordinates of router_AS1 and router_AS2.
1819 This is currently a convention and we may offer to change this convention in the DTD later if needed.
1820 You may have noted that conveniently, a peer named FOO defines an AS named FOO and a router named router_FOO, which is why it works seamlessly with the <b>peer</b> tag.
1823 \subsection pf_routing_howto_choose_wisely Choosing wisely the routing model to use
1826 Choosing wisely the routing model to use can significantly fasten your
1827 simulation/save your time when writing the platform/save tremendous
1828 disk space. Here is the list of available model and their
1829 characteristics (lookup : time to resolve a route):
1831 \li <b>Full</b>: Full routing data (fast, large memory requirements,
1833 \li <b>Floyd</b>: Floyd routing data (slow initialization, fast
1834 lookup, lesser memory requirements, shortest path routing only).
1835 Calculates all routes at once at the beginning.
1836 \li <b>Dijkstra</b>: Dijkstra routing data (fast initialization, slow
1837 lookup, small memory requirements, shortest path routing only).
1838 Calculates a route when necessary.
1839 \li <b>DijkstraCache</b>: Dijkstra routing data (fast initialization,
1840 fast lookup, small memory requirements, shortest path routing
1841 only). Same as Dijkstra, except it handles a cache for latest used
1843 \li <b>None</b>: No routing (usable with Constant network only).
1844 Defines that there is no routes, so if you try to determine a
1845 route without constant network within this AS, SimGrid will raise
1847 \li <b>Vivaldi</b>: Vivaldi routing, so when you want to use coordinates
1848 \li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
1851 \subsection pf_switch I want to describe a switch but there is no switch tag!
1853 Actually we did not include switch tag. But when you're trying to
1854 simulate a switch, assuming
1855 fluid bandwidth models are used (which SimGrid uses by default unless
1856 ns-3 or constant network models are activated), the limiting factor is
1857 switch backplane bandwidth. So, essentially, at least from
1858 the simulation perspective, a switch is similar to a
1859 link: some device that is traversed by flows and with some latency and
1860 so,e maximum bandwidth. Thus, you can simply simulate a switch as a
1862 can be connected to this "switch", which is then included in routes just
1866 \subsection pf_multicabinets I want to describe multi-cabinets clusters!
1868 You have several possibilities, as usual when modeling things. If your
1869 cabinets are homogeneous and the intercabinet network negligible for
1870 your study, you should just create a larger cluster with all hosts at
1873 In the rare case where your hosts are not homogeneous between the
1874 cabinets, you can create your cluster completely manually. For that,
1875 create an As using the Cluster routing, and then use one
1876 <cabinet> for each cabinet. This cabinet tag can only be used an
1877 As using the Cluster routing schema, and creating
1879 Be warned that creating a cluster manually from the XML with
1880 <cabinet>, <backbone> and friends is rather tedious. The
1881 easiest way to retrieve some control of your model without diving into
1882 the <cluster> internals is certainly to create one separate
1883 <cluster> per cabinet and interconnect them together. This is
1884 what we did in the G5K example platform for the Graphen cluster.
1886 \subsection pf_platform_multipath I want to express multipath routing in platform files!
1888 It is unfortunately impossible to express the fact that there is more
1889 than one routing path between two given hosts. Let's consider the
1890 following platform file:
1893 <route src="A" dst="B">
1896 <route src="B" dst="C">
1899 <route src="A" dst="C">
1904 Although it is perfectly valid, it does not mean that data traveling
1905 from A to C can either go directly (using link 3) or through B (using
1906 links 1 and 2). It simply means that the routing on the graph is not
1907 trivial, and that data do not following the shortest path in number of
1908 hops on this graph. Another way to say it is that there is no implicit
1909 in these routing descriptions. The system will only use the routes you
1910 declare (such as <route src="A" dst="C"><link_ctn
1911 id="3"/></route>), without trying to build new routes by aggregating
1914 You are also free to declare platform where the routing is not
1915 symmetrical. For example, add the following to the previous file:
1918 <route src="C" dst="A">
1924 This makes sure that data from C to A go through B where data from A
1925 to C go directly. Don't worry about realism of such settings since
1926 we've seen ways more weird situation in real settings (in fact, that's
1927 the realism of very regular platforms which is questionable, but
1928 that's another story).
