1 /*! \page platform Describing the virtual platform
5 As @ref starting_components "explained in the introduction," any
6 SimGrid study must entail the description of the platform on which you
7 want to simulate your application. You have to describe **each element
8 of your platform**, such as computing hosts, clusters, each disks,
9 links, etc. You must also define the **routing on your platform**, ie
10 which path is taken between two hosts. Finally, you may also describe
11 an **experimental scenario**, with qualitative changes (e.g.,
12 bandwidth changes representing an external load) and qualitative
13 changes (representing how some elements fail and restart over time).
15 You should really separate your application from the platform
16 description, as it will ease your experimental campain afterward.
17 Mixing them is seen as a really bad experimental practice. The easiest
18 to enforce this split is to put the platform description in a XML
19 file. Many example platforms are provided in the archive, and this
20 page gives all needed details to write such files, as well as some
21 hints and tricks about describing your platform.
23 On the other side, XML is sometimes not expressive enough for some
24 platforms, in particular large platforms exhibiting repetitive
25 patterns that are not simply expressed in XML. In practice, many
26 users end up generating their XML platform files from some sort of
27 scripts. It is probably preferable to rewrite your XML @ref
28 platform_lua "platform using the lua scripting language" instead.
29 In the future, it should be possible to describe the platform directly
30 in C++, but this is not possible yet.
32 As usual, SimGrid is a versatile framework, and you should find the
33 way of describing your platform that best fits your experimental
36 \section pf_overview Describing the platform with XML
38 Your platform description should follow the specification presented in
39 the [simgrid.dtd](http://simgrid.gforge.inria.fr/simgrid/simgrid.dtd)
40 DTD file. The same DTD is used for both the platform and deployment
43 From time to time, this DTD evolves to introduce possibly
44 backward-incompatible changes. That is why each platform desciption is
45 enclosed within a @c platform tag, that have a @c version attribute.
46 The current version is <b>4.1</b>. The @c simgrid_update_xml program can
47 upgrade most of the past platform files to the recent formalism.
49 \section pf_first_example First Platform Example
51 Here is a very simple platform file, containing 3 resources (two hosts
52 and one link), and explicitly giving the route between the hosts.
56 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid/simgrid.dtd">
57 <platform version="4.1">
58 <zone id="first zone" routing="Full">
59 <!-- the resources -->
60 <host id="host1" speed="1Mf"/>
61 <host id="host2" speed="2Mf"/>
62 <link id="link1" bandwidth="125MBps" latency="100us"/>
63 <!-- the routing: specify how the hosts are interconnected -->
64 <route src="host1" dst="host2">
65 <link_ctn id="link1"/>
71 As we said, the englobing @ref pf_overview "<platform>" tag is
72 used to specify the dtd version used for this file.
74 Then, every resource (specified with @ref pf_tag_host, @ref
75 pf_tag_link or others) must be located within a given **networking
76 zone**. Each zone is in charge of the routing between its
77 resources. It means that when an host wants to communicate with
78 another host of the same zone, it is the zone's duty to find the list
79 of links that are involved in the communication. Here, since the @ref
80 pf_tag_zone tag has **Full** as a **routing attribute**, all routes
81 must be explicitely given using the @ref pf_tag_route and @ref
82 pf_tag_linkctn tags (this @ref pf_rm "routing model" is both simple
83 and inefficient :) It is OK to not specify the route between two
84 hosts, as long as the processes located on them never try to
87 A zone can contain several zones itself, leading to a hierarchical
88 decomposition of the platform. This can be more efficient (as the
89 inter-zone routing gets factorized with @ref pf_tag_zoneroute), and
90 allows to have more than one routing model in your platform. For
91 example, you could have a coordinate-based routing for the WAN parts
92 of your platforms, a full routing within each datacenter, and a highly
93 optimized routing within each cluster of the datacenter. In this
94 case, determining the route between two given hosts gets @ref
95 routing_basics "somewhat more complex" but SimGrid still computes
96 these routes for you in a time- and space-efficient manner.
97 Here is an illustration of these concepts:
99 
101 Circles represent processing units and squares represent network
102 routers. Bold lines represent communication links. The zone "AS2"
103 models the core of a national network interconnecting a small flat
104 cluster (AS4) and a larger hierarchical cluster (AS5), a subset of a
105 LAN (AS6), and a set of peers scattered around the world (AS7).
107 \section pf_pftags Resource description
109 \subsection pf_tag_zone <zone>
111 Before SimGrid v3.16, networking zones used to be called Autonomous
112 Systems, but this was misleading as zones may include other zones in a
113 hierarchical manner. If you find any remaining reference to ASes,
114 please report this as a bug.
116 Attribute | Value | Description
117 ----------- | ------------------------------------------------- | ----------------------------------------------
118 id | String (mandatory) | The identifier of this zone (must be unique)
119 routing | One of the existing routing algorithm (mandatory) | See Section \ref pf_rm for details.
123 <AS id="AS0" routing="Full">
124 <host id="host1" speed="1000000000"/>
125 <host id="host2" speed="1000000000"/>
126 <link id="link1" bandwidth="125000000" latency="0.000100"/>
127 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
131 In this example, AS0 contains two hosts (host1 and host2). The route
132 between the hosts goes through link1.
134 \subsection pf_Cr Computing resources: hosts, clusters and peers.
136 \subsubsection pf_tag_host <host>
138 A <b>host</b> represents a computer/node card. Every host is able to execute
139 code and it can send and receive data to/from other hosts. Most importantly,
140 a host can contain more than 1 core.
144 Attribute name | Mandatory | Values | Description
145 --------------- | --------- | ------ | -----------
146 id | yes | String | The identifier of the host. facilitates referring to this AS.
147 speed | yes | double (must be > 0.0) | Computational power of every core of this host in FLOPS. Must be larger than 0.0.
148 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.
149 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.
150 state_file | no | string | Same mechanism as availability_file.<br /> <b>Note:</b> The filename must be specified with your system's format.
151 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.
152 pstate | no | double (Default: 0.0) | FIXME: Not yet documented.
154 ### Possible children: ###
156 Tag name | Description | Documentation
157 ------------ | ----------- | -------------
158 \<mount/\> | Defines mounting points between some storage resource and the host. | \ref pf_storage_entity_mount
159 \<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
164 <host id="host1" speed="1000000000"/>
165 <host id="host2" speed="1000000000">
166 <prop id="color" value="blue"/>
167 <prop id="rendershape" value="square"/>
172 \anchor pf_host_dynamism
173 ### Expressing dynamism ###
175 SimGrid provides mechanisms to change a hosts' availability over
176 time, using the ``availability_file`` attribute to the ``\<host\>`` tag
177 and a separate text file whose syntax is exemplified below.
179 #### Adding a trace file ####
182 <platform version="4">
183 <host id="bob" speed="500Gf" availability_file="bob.trace" />
187 #### Example of "bob.trace" file ####
196 Let us begin to explain this example by looking at line 2. (Line 1 will become clear soon).
197 The first column describes points in time, in this case, time 0. The second column
198 describes the relative amount of power this host is able to deliver (relative
199 to the maximum performance specified in the ``\<host\>`` tag). (Clearly, the
200 second column needs to contain values that are not smaller than 0 and not larger than 1).
201 In this example, our host will deliver 500 Mflop/s at time 0, as 500 Mflop/s is the
202 maximum performance of this host. At time 11.0, it will
203 deliver half of its maximum performance, i.e., 250 Mflop/s until time 20.0 when it will
204 will start delivering 80\% of its power. In this example, this amounts to 400 Mflop/s.
206 Since the periodicity in line 1 was set to be 1.0, i.e., 1 timestep, this host will
207 continue to provide 500 Mflop/s from time 21. From time 32 it will provide 250 MFlop/s and so on.
209 ### Changing initial state ###
211 It is also possible to specify whether the host is up or down by setting the
212 ``state`` attribute to either <b>ON</b> (default value) or <b>OFF</b>.
214 #### Example: Expliciting the default value "ON" ####
217 <platform version="4">
218 <host id="bob" speed="500Gf" state="ON" />
222 If you want this host to be unavailable, simply substitute ON with OFF.
224 \anchor pf_host_churn
225 ### Expressing churn ###
227 To express the fact that a host can change state over time (as in P2P
228 systems, for instance), it is possible to use a file describing the time
229 at which the host is turned on or off. An example of the content
230 of such a file is presented below.
232 #### Adding a state file ####
235 <platform version="4">
236 <host id="bob" power="500Gf" state_file="bob.fail" />
240 #### Example of "bob.fail" file ####
248 A zero value means <b>down</b> (i.e., OFF) while a positive one means <b>up and
249 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
250 turned off and at time 2.0, it is turned on again until time 12 (2 plus the
251 periodicity 10). It will be turned off again at time 13.0 until time 23.0, and
255 \subsubsection pf_cluster <cluster>
257 ``<cluster />`` represents a machine-cluster. It is most commonly used
258 when one wants to define many hosts and a network quickly. Technically,
259 ``cluster`` is a meta-tag: <b>from the inner SimGrid point of
260 view, a cluster is an AS where some optimized routing is defined</b>.
261 The default inner organization of the cluster is as follow:
267 ____________|__________|_____________ backbone
269 l0| l1| l2| l97| l96 | | l99
275 Here, a set of <b>host</b>s is defined. Each of them has a <b>link</b>
276 to a central backbone (backbone is a link itself, as a link can
277 be used to represent a switch, see the switch / link section
278 below for more details about it). A <b>router</b> allows to connect a
279 <b>cluster</b> to the outside world. Internally,
280 SimGrid treats a cluster as an AS containing all hosts: the router is the default
281 gateway for the cluster.
283 There is an alternative organization, which is as follows:
297 The principle is the same, except that there is no backbone. This representation
298 can be obtained easily: just do not set the bb_* attributes.
301 Attribute name | Mandatory | Values | Description
302 --------------- | --------- | ------ | -----------
303 id | yes | string | The identifier of the cluster. Facilitates referring to this cluster.
304 prefix | yes | string | Each node of the cluster has to have a name. This name will be prefixed with this prefix.
305 suffix | yes | string | Each node of the cluster will be suffixed with this suffix
306 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.
307 speed | yes | int | Same as the ``speed`` attribute of the ``\<host\>`` tag.
308 core | no | int (default: 1) | Same as the ``core`` attribute of the ``\<host\>`` tag.
309 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
310 lat | yes | int | Latency for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
311 sharing_policy | no | string | Sharing policy for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
312 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>).
313 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>).
314 bb_sharing_policy | no | string | Sharing policy for the backbone (if any). See <b>link</b> section for syntax/details.
315 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>
316 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".
317 loopback_lat | no | int | Latency for loopback (if any). See <b>link</b> section for syntax/details. See loopback_bw for more info.
318 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.
319 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. Please refer to the specific documentation for \ref simgrid::kernel::routing::FatTreeZone "FatTree NetZone", \ref simgrid::kernel::routing::DragonflyZone "Dragonfly NetZone".
322 the router name is defined as the resulting String in the following
326 router_name = prefix + clusterId + "_router" + suffix;
330 #### Cluster example ####
332 Consider the following two (and independent) uses of the ``cluster`` tag:
335 <cluster id="my_cluster_1" prefix="" suffix="" radical="0-262144"
336 speed="1e9" bw="125e6" lat="5E-5"/>
338 <cluster id="my_cluster_2" prefix="c-" suffix=".me" radical="0-99"
339 speed="1e9" bw="125e6" lat="5E-5"
340 bb_bw="2.25e9" bb_lat="5E-4"/>
343 The second example creates one router and 100 machines with the following names:
345 c-my_cluster_2_router.me
353 \subsubsection pf_cabinet <cabinet>
356 This tag is only available when the routing mode of the AS
357 is set to ``Cluster``.
359 The ``<cabinet />`` tag is, like the \ref pf_cluster "<cluster>" tag,
360 a meta-tag. This means that it is simply a shortcut for creating a set of (homogenous) hosts and links quickly;
361 unsurprisingly, this tag was introduced to setup cabinets in data centers quickly. Unlike
362 <cluster>, however, the <cabinet> assumes that you create the backbone
363 and routers yourself; see our examples below.
367 Attribute name | Mandatory | Values | Description
368 --------------- | --------- | ------ | -----------
369 id | yes | string | The identifier of the cabinet. Facilitates referring to this cluster.
370 prefix | yes | string | Each node of the cabinet has to have a name. This name will be prefixed with this prefix.
371 suffix | yes | string | Each node of the cabinet will be suffixed with this suffix
372 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.
373 speed | yes | int | Same as the ``speed`` attribute of the \ref pf_tag_host "<host>" tag.
374 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
375 lat | yes | int | Latency for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
378 Please note that as of now, it is impossible to change attributes such as,
379 amount of cores (always set to 1), the initial state of hosts/links
380 (always set to ON), the sharing policy of the links (always set to \ref pf_sharing_policy_fullduplex "FULLDUPLEX").
384 The following example was taken from ``examples/platforms/meta_cluster.xml`` and
385 shows how to use the cabinet tag.
388 <AS id="my_cluster1" routing="Cluster">
389 <cabinet id="cabinet1" prefix="host-" suffix=".cluster1"
390 speed="1Gf" bw="125MBps" lat="100us" radical="1-10"/>
391 <cabinet id="cabinet2" prefix="host-" suffix=".cluster1"
392 speed="1Gf" bw="125MBps" lat="100us" radical="11-20"/>
393 <cabinet id="cabinet3" prefix="host-" suffix=".cluster1"
394 speed="1Gf" bw="125MBps" lat="100us" radical="21-30"/>
396 <backbone id="backbone1" bandwidth="2.25GBps" latency="500us"/>
401 Please note that you must specify the \ref pf_backbone "<backbone>"
402 tag by yourself; this is not done automatically and there are no checks
403 that ensure this backbone was defined.
405 The hosts generated in the above example are named host-1.cluster, host-2.cluster1
409 \subsubsection pf_peer \<peer\> (Vivaldi netzones only)
411 This tag represents a peer, as in Peer-to-Peer (P2P) networks. This
412 can only be used in Vivaldi NetZones. It creates the following
413 resources to the NetZone:
416 \li Two links: One for download and one for upload. This is
417 convenient to use and simulate stuff under the last mile model (e.g., ADSL peers).
418 \li It connects the two links to the host
422 Attribute name | Mandatory | Values | Description
423 --------------- | --------- | ------ | -----------
424 id | yes | string | The identifier of the peer. Facilitates referring to this peer.
425 speed | yes | int | See the description of the ``host`` tag for this attribute
426 bw_in | yes | int | Bandwidth of the private downstream link
427 bw_out | yes | int | Bandwidth of the private upstream link
428 coordinates | no | string | Coordinates of the gateway for this peer. Example value: 12.8 14.4 6.4
429 sharing_policy | no | SHARED\|FULLDUPLEX (default: FULLDUPLEX) | Sharing policy for links. See <b>link</b> description for details.
430 availability_file| no | string | Availability file for the peer. Same as host availability file. See <b>host</b> description for details.
431 state_file | no | string | State file for the peer. Same as host state file. See <b>host</b> description for details.
434 The communication latency between an host A=(xA,yA,zA) and an host
435 B=(xB,yB,zB) is computed as follows:
437 latency = sqrt( (xA-xB)² + (yA-yB)² ) + zA + zB
439 See the documentation of simgrid::kernel::routing::VivaldiZone for
440 details on how the latency is computed from the coordinate, and on the
441 the up and down bandwidth are used.
443 \subsection pf_ne Network equipments: links and routers
445 There are two tags at all times available to represent network entities and
446 several other tags that are available only in certain contexts.
447 1. ``<link>``: Represents a entity that has a limited bandwidth, a
448 latency, and that can be shared according to TCP way to share this
451 The concept of links in SimGrid may not be intuitive, as links are not
452 limited to connecting (exactly) two entities; in fact, you can have more than
453 two equipments connected to it. (In graph theoretical terms: A link in
454 SimGrid is not an edge, but a hyperedge)
456 2. ``<router/>``: Represents an entity that a message can be routed
457 to, but that is unable to execute any code. In SimGrid, routers have also
458 no impact on the performance: Routers do not limit any bandwidth nor
459 do they increase latency. As a matter of fact, routers are (almost) ignored
460 by the simulator when the simulation has begun.
462 3. ``<backbone/>``: This tag is only available when the containing AS is
463 used as a cluster (i.e., mode="Cluster")
466 If you want to represent an entity like a switch, you must use ``<link>`` (see section). Routers are used
467 to run some routing algorithm and determine routes (see Section \ref pf_routing for details).
469 \subsubsection pf_router <router/>
471 As said before, <b>router</b> is used only to give some information
472 for routing algorithms. So, it does not have any attributes except :
476 Attribute name | Mandatory | Values | Description
477 --------------- | --------- | ------ | -----------
478 id | yes | string | The identifier of the router to be used when referring to it.
479 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.
484 <router id="gw_dc1_horizdist"/>
487 \subsubsection pf_tag_link <link>
489 Network links can represent one-hop network connections. They are
490 characterized by their id and their bandwidth; links can (but may not) be subject
495 Attribute name | Mandatory | Values | Description
496 --------------- | --------- | ------ | -----------
497 id | yes | string | The identifier of the link to be used when referring to it.
498 bandwidth | yes | int | Maximum bandwidth for this link, given in bytes/s
499 latency | no | double (default: 0.0) | Latency for this link.
500 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.
501 state | no | ON\|OFF (default: ON) | Allows you to to turn this link on or off (working / not working)
502 bandwidth_file | no | string | Allows you to use a file as input for bandwidth.
503 latency_file | no | string | Allows you to use a file as input for latency.
504 state_file | no | string | Allows you to use a file as input for states.
507 #### Possible shortcuts for ``latency`` ####
509 When using the latency attribute, you can specify the latency by using the scientific
510 notation or by using common abbreviations. For instance, the following three tags
514 <link id="LINK1" bandwidth="125000000" latency="5E-6"/>
515 <link id="LINK1" bandwidth="125000000" latency="5us"/>
516 <link id="LINK1" bandwidth="125000000" latency="0.000005"/>
519 Here, the second tag uses "us", meaning "microseconds". Other shortcuts are:
521 Name | Abbreviation | Time (in seconds)
522 ---- | ------------ | -----------------
523 Week | w | 7 * 24 * 60 * 60
524 Day | d | 24 * 60 * 60
528 Millisecond | ms | 0.001 = 10^(-3)
529 Microsecond | us | 0.000001 = 10^(-6)
530 Nanosecond | ns | 0.000000001 = 10^(-9)
531 Picosecond | ps | 0.000000000001 = 10^(-12)
533 #### Sharing policy ####
535 \anchor sharing_policy_shared
536 By default a network link is \b SHARED, i.e., if two or more data flows go
537 through a link, the bandwidth is shared fairly among all data flows. This
538 is similar to the sharing policy TCP uses.
540 \anchor pf_sharing_policy_fatpipe
541 On the other hand, if a link is defined as a \b FATPIPE,
542 each flow going through this link will be provided with the complete bandwidth,
543 i.e., no sharing occurs and the bandwidth is only limiting each flow individually.
544 Please note that this is really on a per-flow basis, not only on a per-host basis!
545 The complete bandwidth provided by this link in this mode
546 is ``number_of_flows*bandwidth``, with at most ``bandwidth`` being available per flow.
548 Using the FATPIPE mode allows to model backbones that won't affect performance
551 \anchor pf_sharing_policy_fullduplex
552 The last mode available is \b FULLDUPLEX. This means that SimGrid will
553 automatically generate two links (one carrying the suffix _UP and the other the
554 suffix _DOWN) for each ``<link>`` tag. This models situations when the direction
555 of traffic is important.
558 Transfers from one side to the other will interact similarly as
559 TCP when ACK returning packets circulate on the other direction. More
560 discussion about it is available in the description of link_ctn description.
562 In other words: The SHARED policy defines a physical limit for the bandwidth.
563 The FATPIPE mode defines a limit for each application,
564 with no upper total limit.
567 Tip: By using the FATPIPE mode, you can model big backbones that
568 won't affect performance (except latency).
573 <link id="SWITCH" bandwidth="125000000" latency="5E-5" sharing_policy="FATPIPE" />
576 #### Expressing dynamism and failures ####
578 Similar to hosts, it is possible to declare links whose state, bandwidth
579 or latency changes over time (see Section \ref pf_host_dynamism for details).
581 In the case of network links, the ``bandwidth`` and ``latency`` attributes are
582 replaced by the ``bandwidth_file`` and ``latency_file`` attributes.
583 The following XML snippet demonstrates how to use this feature in the platform
584 file. The structure of the files "link1.bw" and "link1.lat" is shown below.
587 <link id="LINK1" state_file="link1.fail" bandwidth="80000000" latency=".0001" bandwidth_file="link1.bw" latency_file="link1.lat" />
591 Even if the syntax is the same, the semantic of bandwidth and latency
592 trace files differs from that of host availability files. For bandwidth and
593 latency, the corresponding files do not
594 express availability as a fraction of the available capacity but directly in
595 bytes per seconds for the bandwidth and in seconds for the latency. This is
596 because most tools allowing to capture traces on real platforms (such as NWS)
597 express their results this way.
599 ##### Example of "link1.bw" file #####
607 In this example, the bandwidth changes repeatedly, with all changes
608 being repeated every 12 seconds.
610 At the beginning of the the simulation, the link's bandwidth is 80,000,000
611 B/s (i.e., 80 Mb/s); this value was defined in the XML snippet above.
612 After four seconds, it drops to 40 Mb/s (line 2), and climbs
613 back to 60 Mb/s after another 4 seconds (line 3). The value does not change any
614 more until the end of the period, that is, after 12 seconds have been simulated).
615 At this point, periodicity kicks in and this behavior is repeated: Seconds
616 12-16 will experience 80 Mb/s, 16-20 40 Mb/s etc.).
618 ##### Example of "link1.lat" file #####
627 In this example, the latency varies with a period of 5 seconds.
628 In the xml snippet above, the latency is initialized to be 0.0001s (100µs). This
629 value will be kept during the first second, since the latency_file contains
630 changes to this value at second one, two and three.
631 At second one, the value will be 0.001, i.e., 1ms. One second later it will
632 be adjusted to 0.01 (or 10ms) and one second later it will be set again to 1ms. The
633 value will not change until second 5, when the periodicity defined in line 1
634 kicks in. It then loops back, starting at 100µs (the initial value) for one second.
637 #### The ``<prop/>`` tag ####
639 Similar to the ``<host>`` tag, a link may also contain the ``<prop/>`` tag; see the host
640 documentation (Section \ref pf_host) for an example.
643 \subsubsection pf_backbone <backbone/>
646 This tag is <b>only available</b> when the containing AS uses the "Cluster" routing mode!
648 Using this tag, you can designate an already existing link to be a backbone.
650 Attribute name | Mandatory | Values | Description
651 --------------- | --------- | ------ | -----------
652 id | yes | string | Name of the link that is supposed to act as a backbone.
654 \subsection pf_storage Storage
657 This is a prototype version that should evolve quickly, hence this
658 is just some doc valuable only at the time of writing.
659 This section describes the storage management under SimGrid ; nowadays
660 it's only usable with MSG. It relies basically on linux-like concepts.
661 You also may want to have a look to its corresponding section in
662 @ref msg_file ; access functions are organized as a POSIX-like
665 \subsubsection pf_sto_conc Storage - Main Concepts
667 The storage facilities implemented in SimGrid help to model (and account for)
668 storage devices, such as tapes, hard-drives, CD or DVD devices etc.
669 A typical situation is depicted in the figure below:
671 \image html ./webcruft/storage_sample_scenario.png
672 \image latex ./webcruft/storage_sample_scenario.png "storage_sample_scenario" width=\textwidth
674 In this figure, two hosts called Bob and Alice are interconnected via a network
675 and each host is physically attached to a disk; it is not only possible for each host to
676 mount the disk they are attached to directly, but they can also mount disks
677 that are in a remote location. In this example, Bob mounts Alice's disk remotely
678 and accesses the storage via the network.
680 SimGrid provides 3 different entities that can be used to model setups
681 that include storage facilities:
683 Entity name | Description
684 --------------- | -----------
685 \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).
686 \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.
687 \ref pf_storage_entity_mount "mount" | Must be wrapped by a \ref pf_tag_host tag; declares which storage(s) this host has mounted and where (i.e., the mountpoint).
690 \anchor pf_storage_content_file
691 ### %Storage Content File ###
693 In order to assess exactly how much time is spent reading from the storage,
694 SimGrid needs to know what is stored on the storage device (identified by distinct (file-)name, like in a file system)
695 and what size this content has.
698 The content file is never changed by the simulation; it is parsed once
699 per simulation and kept in memory afterwards. When the content of the
700 storage changes, only the internal SimGrid data structures change.
702 \anchor pf_storage_content_file_structure
703 #### Structure of a %Storage Content File ####
705 Here is an excerpt from two storage content file; if you want to see the whole file, check
706 the file ``examples/platforms/content/storage_content.txt`` that comes with the
709 SimGrid essentially supports two different formats: UNIX-style filepaths should
710 follow the well known format:
713 /lib/libsimgrid.so.3.6.2 12710497
717 /bin/simgrid_update_xml 5018
718 /bin/graphicator 66986
719 /bin/simgrid-colorizer 2993
724 Windows filepaths, unsurprisingly, use the windows style:
727 \Windows\avastSS.scr 41664
728 \Windows\bfsvc.exe 75264
729 \Windows\bootstat.dat 67584
730 \Windows\CoreSingleLanguage.xml 31497
732 \Windows\dchcfg64.exe 335464
733 \Windows\dcmdev64.exe 93288
737 The different file formats come at a cost; in version 3.12 (and most likely
738 in later versions, too), copying files from windows-style storages to unix-style
739 storages (and vice versa) is not supported.
741 \anchor pf_storage_content_file_create
742 #### Generate a %Storage Content File ####
744 If you want to generate a storage content file based on your own filesystem (or at least a filesystem you have access to),
745 try running this command (works only on unix systems):
748 find . -type f -exec ls -1s --block=1 {} \; 2>/dev/null | awk '{ print $2 " " $1}' > ./content.txt
751 \subsubsection pf_storage_entities The Storage Entities
753 These are the entities that you can use in your platform files to include
754 storage in your model. See also the list of our \ref pf_storage_example_files "example files";
755 these might also help you to get started.
757 \anchor pf_storage_entity_storage_type
758 #### \<storage_type\> ####
760 Attribute name | Mandatory | Values | Description
761 --------------- | --------- | ------ | -----------
762 id | yes | string | Identifier of this storage_type; used when referring to it
763 model | yes | string | For reasons of future backwards compatibility only; specifies the name of the model for the storage that should be used
764 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)
765 content | yes | string | Path to a \ref pf_storage_content_file "Storage Content File" on your system. This file must exist.
766 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"!)
768 This tag must contain some predefined model properties, specified via the <model_prop> tag. Here is a list,
769 see below for an example:
771 Property id | Mandatory | Values | Description
772 --------------- | --------- | ------ | -----------
773 Bwrite | yes | string | Bandwidth for write access; in B/s (but you can also specify e.g. "30MBps")
774 Bread | yes | string | Bandwidth for read access; in B/s (but you can also specify e.g. "30MBps")
775 Bconnexion | yes | string | Throughput (of the storage connector) in B/s.
778 A storage_type can also contain the <b><prop></b> tag. The <prop> tag allows you
779 to associate additional information to this <storage_type> and follows the
780 attribute/value schema; see the example below. You may want to use it to give information to
781 the tool you use for rendering your simulation, for example.
783 Here is a complete example for the ``storage_type`` tag:
785 <storage_type id="single_HDD" model="linear_no_lat" size="4000" content_type="txt_unix">
786 <model_prop id="Bwrite" value="30MBps" />
787 <model_prop id="Bread" value="100MBps" />
788 <model_prop id="Bconnection" value="150MBps" />
789 <prop id="Brand" value="Western Digital" />
793 \anchor pf_storage_entity_storage
794 #### <storage> ####
796 ``storage`` attributes:
798 Attribute name | Mandatory | Values | Description
799 -------------- | --------- | ------ | -----------
800 id | yes | string | Identifier of this ``storage``; used when referring to it
801 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.
802 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)
803 content | no | string | When specified, overwrites the content attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
804 content_type | no | string | When specified, overwrites the content_type attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
806 Here are two examples:
809 <storage id="Disk1" typeId="single_HDD" attach="bob" />
811 <storage id="Disk2" typeId="single_SSD"
812 content="content/win_storage_content.txt"
813 content_type="txt_windows" attach="alice" />
816 The first example is straightforward: A disk is defined and called "Disk1"; it is
817 of type "single_HDD" (shown as an example of \ref pf_storage_entity_storage_type "\<storage_type\>" above) and attached
818 to a host called "bob" (the definition of this host is omitted here).
820 The second storage is called "Disk2", is still of the same type as Disk1 but
821 now specifies a new content file (so the contents will be different from Disk1)
822 and the filesystem uses the windows style; finally, it is attached to a second host,
823 called alice (which is again not defined here).
825 \anchor pf_storage_entity_mount
826 #### <mount> ####
829 | Attribute name | Mandatory | Values | Description |
830 | ---------------- | ----------- | -------- | ------------- |
831 | id | yes | string | Refers to a \ref pf_storage_entity_storage "<storage>" entity that will be mounted on that computer |
832 | name | yes | string | Path/location to/of the logical reference (mount point) of this disk
834 This tag must be enclosed by a \ref pf_tag_host tag. It then specifies where the mountpoint of a given storage device (defined by the ``id`` attribute)
835 is; this location is specified by the ``name`` attribute.
837 Here is a simple example, taken from the file ``examples/platform/storage.xml``:
840 <storage_type id="single_SSD" model="linear_no_lat" size="500GiB">
841 <model_prop id="Bwrite" value="60MBps" />
842 <model_prop id="Bread" value="200MBps" />
843 <model_prop id="Bconnection" value="220MBps" />
846 <storage id="Disk2" typeId="single_SSD"
847 content="content/win_storage_content.txt"
848 content_type="txt_windows" attach="alice" />
849 <storage id="Disk4" typeId="single_SSD"
850 content="content/small_content.txt"
851 content_type="txt_unix" attach="denise"/>
853 <host id="alice" speed="1Gf">
854 <mount storageId="Disk2" name="c:"/>
857 <host id="denise" speed="1Gf">
858 <mount storageId="Disk2" name="c:"/>
859 <mount storageId="Disk4" name="/home"/>
863 This example is quite interesting, as the same device, called "Disk2", is mounted by
864 two hosts at the same time! Note, however, that the host called ``alice`` is actually
865 attached to this storage, as can be seen in the \ref pf_storage_entity_storage "<storage>"
866 tag. This means that ``denise`` must access this storage through the network, but SimGrid automatically takes
867 care of that for you.
869 Furthermore, this example shows that ``denise`` has mounted two storages with different
870 filesystem types (unix and windows). In general, a host can mount as many storage devices as
874 Again, the difference between ``attach`` and ``mount`` is simply that
875 an attached storage is always physically inside (or connected to) that machine;
876 for instance, a USB stick is attached to one and only one machine (where it's plugged-in)
877 but it can only be mounted on others, as mounted storage can also be a remote location.
879 ###### Example files #####
881 \verbinclude example_filelist_xmltag_mount
883 \subsubsection pf_storage_example_files Example files
885 Several examples were already discussed above; if you're interested in full examples,
886 check the the following platforms:
888 1. ``examples/platforms/storage.xml``
889 2. ``examples/platforms/remote_io.xml``
891 If you're looking for some examplary C code, you may find the source code
892 available in the directory ``examples/msg/io/`` useful.
894 \subsubsection pf_storage_examples_modelling Modelling different situations
896 The storage functionality of SimGrid is type-agnostic, that is, the implementation
897 does not presume any type of storage, such as HDDs/SSDs, RAM,
898 CD/DVD devices, USB sticks etc.
900 This allows the user to apply the simulator for a wide variety of scenarios; one
901 common scenario would be the access of remote RAM.
903 #### Modelling the access of remote RAM ####
905 How can this be achieved in SimGrid? Let's assume we have a setup where three hosts
906 (HostA, HostB, HostC) need to access remote RAM:
916 An easy way to model this scenario is to setup and define the RAM via the
917 \ref pf_storage_entity_storage "storage" and \ref pf_storage_entity_storage_type "storage type"
918 entities and attach it to a remote dummy host; then, every host can have their own links
919 to this host (modelling for instance certain scenarios, such as PCIe ...)
924 RAM - Dummy -- Host B
929 Now, if read from this storage, the host that mounts this storage
930 communicates to the dummy host which reads from RAM and
931 sends the information back.
934 \section pf_routing Routing
936 To achieve high performance, the routing tables used within SimGrid are
937 static. This means that routing between two nodes is calculated once
938 and will not change during execution. The SimGrid team chose to use this
939 approach as it is rare to have a real deficiency of a resource;
940 most of the time, a communication fails because the links experience too much
941 congestion and hence, your connection stops before the timeout or
942 because the computer designated to be the destination of that message
945 We also chose to use shortest paths algorithms in order to emulate
946 routing. Doing so is consistent with the reality: [RIP](https://en.wikipedia.org/wiki/Routing_Information_Protocol),
947 [OSPF](https://en.wikipedia.org/wiki/Open_Shortest_Path_First), [BGP](https://en.wikipedia.org/wiki/Border_Gateway_Protocol)
948 are all calculating shortest paths. They do require some time to converge, but
949 eventually, when the routing tables have stabilized, your packets will follow
952 \subsection pf_rm Routing models
954 For each AS, you must define explicitly which routing model will
955 be used. There are 3 different categories for routing models:
957 1. \ref pf_routing_model_shortest_path "Shortest-path" based models: SimGrid calculates shortest
958 paths and manages them. Behaves more or less like most real life
960 2. \ref pf_routing_model_manual "Manually-entered" route models: you have to define all routes
961 manually in the platform description file; this can become
962 tedious very quickly, as it is very verbose.
963 Consistent with some manually managed real life routing.
964 3. \ref pf_routing_model_simple "Simple/fast models": those models offer fast, low memory routing
965 algorithms. You should consider to use this type of model if
966 you can make some assumptions about your AS.
967 Routing in this case is more or less ignored.
969 \subsubsection pf_raf The router affair
971 Using routers becomes mandatory when using shortest-path based
972 models or when using the bindings to the ns-3 packet-level
973 simulator instead of the native analytical network model implemented
976 For graph-based shortest path algorithms, routers are mandatory, because these
977 algorithms require a graph as input and so we need to have source and
978 destination for each edge.
980 Routers are naturally an important concept ns-3 since the
981 way routers run the packet routing algorithms is actually simulated.
982 SimGrid's analytical models however simply aggregate the routing time
983 with the transfer time.
985 So why did we incorporate routers in SimGrid? Rebuilding a graph representation
986 only from the route information turns out to be a very difficult task, because
987 of the missing information about how routes intersect. That is why we
988 introduced routers, which are simply used to express these intersection points.
989 It is important to understand that routers are only used to provide topological
992 To express this topological information, a <b>route</b> has to be
993 defined in order to declare which link is connected to a router.
996 \subsubsection pf_routing_model_shortest_path Shortest-path based models
998 The following table shows all the models that compute routes using
999 shortest-paths algorithms are currently available in SimGrid. More detail on how
1000 to choose the best routing model is given in the Section called \"\ref pf_routing_howto_choose_wisely\".
1002 | Name | Description |
1003 | --------------------------------------------------- | -------------------------------------------------------------------------- |
1004 | \ref pf_routing_model_floyd "Floyd" | Floyd routing data. Pre-calculates all routes once |
1005 | \ref pf_routing_model_dijkstra "Dijkstra" | Dijkstra routing data. Calculates routes only when needed |
1006 | \ref pf_routing_model_dijkstracache "DijkstraCache" | Dijkstra routing data. Handles some cache for already calculated routes. |
1008 All those shortest-path models are instanciated in the same way and are
1009 completely interchangeable. Here are some examples:
1011 \anchor pf_routing_model_floyd
1016 <AS id="AS0" routing="Floyd">
1018 <cluster id="my_cluster_1" prefix="c-" suffix=""
1019 radical="0-1" speed="1000000000" bw="125000000" lat="5E-5"
1020 router_id="router1"/>
1022 <AS id="AS1" routing="None">
1023 <host id="host1" speed="1000000000"/>
1026 <link id="link1" bandwidth="100000" latency="0.01"/>
1028 <ASroute src="my_cluster_1" dst="AS1"
1031 <link_ctn id="link1"/>
1037 ASroute given at the end gives a topological information: link1 is
1038 between router1 and host1.
1040 #### Example platform files ####
1042 This is an automatically generated list of example files that use the Floyd
1043 routing model (the path is given relative to SimGrid's source directory)
1045 \verbinclude example_filelist_routing_floyd
1047 \anchor pf_routing_model_dijkstra
1050 #### Example platform files ####
1052 This is an automatically generated list of example files that use the Dijkstra
1053 routing model (the path is given relative to SimGrid's source directory)
1055 \verbinclude example_filelist_routing_dijkstra
1059 <AS id="AS_2" routing="Dijkstra">
1060 <host id="AS_2_host1" speed="1000000000"/>
1061 <host id="AS_2_host2" speed="1000000000"/>
1062 <host id="AS_2_host3" speed="1000000000"/>
1063 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1064 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1065 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1066 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1067 <router id="central_router"/>
1068 <router id="AS_2_gateway"/>
1069 <!-- routes providing topological information -->
1070 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1071 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1072 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1073 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1077 \anchor pf_routing_model_dijkstracache
1078 ### DijkstraCache ###
1080 DijkstraCache example:
1082 <AS id="AS_2" routing="DijkstraCache">
1083 <host id="AS_2_host1" speed="1000000000"/>
1085 (platform unchanged compared to upper example)
1088 #### Example platform files ####
1090 This is an automatically generated list of example files that use the DijkstraCache
1091 routing model (the path is given relative to SimGrid's source directory):
1093 Editor's note: At the time of writing, no platform file used this routing model - so
1094 if there are no example files listed here, this is likely to be correct.
1096 \verbinclude example_filelist_routing_dijkstra_cache
1098 \subsubsection pf_routing_model_manual Manually-entered route models
1100 | Name | Description |
1101 | ---------------------------------- | ------------------------------------------------------------------------------ |
1102 | \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. |
1104 \anchor pf_routing_model_full
1109 <AS id="AS0" routing="Full">
1110 <host id="host1" speed="1000000000"/>
1111 <host id="host2" speed="1000000000"/>
1112 <link id="link1" bandwidth="125000000" latency="0.000100"/>
1113 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
1117 #### Example platform files ####
1119 This is an automatically generated list of example files that use the Full
1120 routing model (the path is given relative to SimGrid's source directory):
1122 \verbinclude example_filelist_routing_full
1124 \subsubsection pf_routing_model_simple Simple/fast models
1126 | Name | Description |
1127 | ---------------------------------------- | ------------------------------------------------------------------------------ |
1128 | \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. |
1129 | \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. |
1130 | \ref pf_routing_model_vivaldi "Vivaldi" | Perfect when you want to use coordinates. Also see the corresponding \ref pf_P2P_tags "P2P section" below. |
1132 \anchor pf_routing_model_cluster
1136 In this mode, the \ref pf_cabinet "<cabinet/>" tag is available.
1138 #### Example platform files ####
1140 This is an automatically generated list of example files that use the Cluster
1141 routing model (the path is given relative to SimGrid's source directory):
1143 \verbinclude example_filelist_routing_cluster
1145 \anchor pf_routing_model_none
1148 This model does exactly what it's name advertises: Nothing. There is no routing
1149 available within this model and if you try to communicate within the AS that
1150 uses this model, SimGrid will fail unless you have explicitly activated the
1151 \ref options_model_select_network_constant "Constant Network Model" (this model charges
1152 the same for every single communication). It should
1153 be noted, however, that you can still attach an \ref pf_tag_asroute "ASroute",
1154 as is demonstrated in the example below:
1156 \verbinclude platforms/cluster_and_one_host.xml
1158 #### Example platform files ####
1160 This is an automatically generated list of example files that use the None
1161 routing model (the path is given relative to SimGrid's source directory):
1163 \verbinclude example_filelist_routing_none
1166 \anchor pf_routing_model_vivaldi
1169 For more information on how to use the [Vivaldi Coordinates](https://en.wikipedia.org/wiki/Vivaldi_coordinates),
1170 see also Section \ref pf_P2P_tags "P2P tags".
1172 Note that it is possible to combine the Vivaldi routing model with other routing models;
1173 an example can be found in the file \c examples/platforms/cloud.xml. This
1174 examples models a NetZone using Vivaldi that contains other NetZones that use different
1177 #### Example platform files ####
1179 This is an automatically generated list of example files that use the None
1180 routing model (the path is given relative to SimGrid's source directory):
1182 \verbinclude example_filelist_routing_vivaldi
1185 \subsection ps_dec Defining routes
1187 There are currently four different ways to define routes:
1189 | Name | Description |
1190 | ------------------------------------------------- | ----------------------------------------------------------------------------------- |
1191 | \ref pf_tag_route "route" | Used to define route between host/router |
1192 | \ref pf_tag_zoneroute "zoneRoute" | Used to define route between different zones |
1193 | \ref pf_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. |
1194 | \ref pf_tag_bypassasroute "bypassZoneRoute" | Used in the same way as bypassRoute, but for zones |
1196 Basically all those tags will contain an (ordered) list of references
1197 to link that compose the route you want to define.
1199 Consider the example below:
1202 <route src="Alice" dst="Bob">
1203 <link_ctn id="link1"/>
1204 <link_ctn id="link2"/>
1205 <link_ctn id="link3"/>
1209 The route here from host Alice to Bob will be first link1, then link2,
1210 and finally link3. What about the reverse route? \ref pf_tag_route "Route" and
1211 \ref pf_tag_zoneroute "ASroute" have an optional attribute \c symmetrical, that can
1212 be either \c YES or \c NO. \c YES means that the reverse route is the same
1213 route in the inverse order, and is set to \c YES by default. Note that
1214 this is not the case for bypass*Route, as it is more probable that you
1215 want to bypass only one default route.
1217 For an \ref pf_tag_zoneroute "ASroute", things are just slightly more complicated, as you have
1218 to give the id of the gateway which is inside the AS you want to access ...
1219 So it looks like this:
1222 <ASroute src="AS1" dst="AS2"
1223 gw_src="router1" gw_dst="router2">
1224 <link_ctn id="link1"/>
1228 gw == gateway, so when any message are trying to go from AS1 to AS2,
1229 it means that it must pass through router1 to get out of the AS, then
1230 pass through link1, and get into AS2 by being received by router2.
1231 router1 must belong to AS1 and router2 must belong to AS2.
1233 \subsubsection pf_tag_linkctn <link_ctn>
1235 This entity has only one purpose: Refer to an already existing
1236 \ref pf_tag_link "<link/>" when defining a route, i.e., it
1237 can only occur as a child of \ref pf_tag_route "<route/>"
1239 | Attribute name | Mandatory | Values | Description |
1240 | --------------- | --------- | ------ | ----------- |
1241 | id | yes | String | The identifier of the link that should be added to the route. |
1242 | 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.
1244 #### Example Files ####
1246 This is an automatically generated list of example files that use the \c <link_ctn/>
1247 entity (the path is given relative to SimGrid's source directory):
1249 \verbinclude example_filelist_xmltag_linkctn
1251 \subsubsection pf_tag_zoneroute <zoneRoute>
1253 The purpose of this entity is to define a route between two ASes.
1254 This is mainly useful when you're in the \ref pf_routing_model_full "Full routing model".
1256 #### Attributes ####
1258 | Attribute name | Mandatory | Values | Description |
1259 | --------------- | --------- | ------ | ----------- |
1260 | src | yes | String | The identifier of the source AS |
1261 | dst | yes | String | See the \c src attribute |
1262 | gw_src | yes | String | The gateway that will be used within the src AS; this can be any \ref pf_tag_host "Host" or \ref pf_router "Router" defined within the src AS. |
1263 | gw_dst | yes | String | Same as \c gw_src, but with the dst AS instead. |
1264 | symmetrical | no | YES\|NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1269 <AS id="AS0" routing="Full">
1270 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
1271 radical="0-149" speed="1000000000" bw="125000000" lat="5E-5"
1272 bb_bw="2250000000" bb_lat="5E-4"/>
1274 <cluster id="my_cluster_2" prefix="c-" suffix=".me"
1275 radical="150-299" speed="1000000000" bw="125000000" lat="5E-5"
1276 bb_bw="2250000000" bb_lat="5E-4"/>
1278 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1280 <ASroute src="my_cluster_1" dst="my_cluster_2"
1281 gw_src="c-my_cluster_1_router.me"
1282 gw_dst="c-my_cluster_2_router.me">
1283 <link_ctn id="backbone"/>
1285 <ASroute src="my_cluster_2" dst="my_cluster_1"
1286 gw_src="c-my_cluster_2_router.me"
1287 gw_dst="c-my_cluster_1_router.me">
1288 <link_ctn id="backbone"/>
1293 \subsubsection pf_tag_route <route>
1295 The principle is the same as for
1296 \ref pf_tag_zoneroute "ASroute": The route contains a list of links that
1297 provide a path from \c src to \c dst. Here, \c src and \c dst can both be either a
1298 \ref pf_tag_host "host" or \ref pf_router "router". This is mostly useful for the
1299 \ref pf_routing_model_full "Full routing model" as well as for the
1300 \ref pf_routing_model_shortest_path "shortest-paths" based models (as they require
1301 topological information).
1304 | Attribute name | Mandatory | Values | Description |
1305 | --------------- | --------- | ---------------------- | ----------- |
1306 | src | yes | String | The value given to the source's "id" attribute |
1307 | dst | yes | String | The value given to the destination's "id" attribute. |
1308 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1313 A route in the \ref pf_routing_model_full "Full routing model" could look like this:
1315 <route src="Tremblay" dst="Bourassa">
1316 <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"/>
1320 A route in the \ref pf_routing_model_shortest_path "Shortest-Path routing model" could look like this:
1322 <route src="Tremblay" dst="Bourassa">
1327 You must only have one link in your routes when you're using them to provide
1328 topological information, as the routes here are simply the edges of the
1329 (network-)graph and the employed algorithms need to know which edge connects
1330 which pair of entities.
1332 \subsubsection pf_tag_bypassasroute bypassASroute
1334 As said before, once you choose
1335 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1336 define some of your routes, which will be specific. You may also want
1337 to bypass some routes defined in lower level AS at an upper stage:
1338 <b>bypassASroute</b> is the tag you're looking for. It allows to
1339 bypass routes defined between already defined between AS (if you want
1340 to bypass route for a specific host, you should just use byPassRoute).
1341 The principle is the same as ASroute : <b>bypassASroute</b> contains
1342 list of links that are in the path between src and dst.
1344 #### Attributes ####
1346 | Attribute name | Mandatory | Values | Description |
1347 | --------------- | --------- | ---------------------- | ----------- |
1348 | src | yes | String | The value given to the source AS's "id" attribute |
1349 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1350 | 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 |
1351 | 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|
1352 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1357 <bypassASRoute src="my_cluster_1" dst="my_cluster_2"
1358 gw_src="my_cluster_1_router"
1359 gw_dst="my_cluster_2_router">
1360 <link_ctn id="link_tmp"/>
1364 This example shows that link \c link_tmp (definition not displayed here) directly
1365 connects the router \c my_cluster_1_router in the source cluster to the router
1366 \c my_cluster_2_router in the destination router. Additionally, as the \c symmetrical
1367 attribute was not given, this route is presumed to be symmetrical.
1369 \subsubsection pf_tag_bypassroute bypassRoute
1371 As said before, once you choose
1372 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1373 define some of your routes, which will be specific. You may also want
1374 to bypass some routes defined in lower level AS at an upper stage :
1375 <b>bypassRoute</b> is the tag you're looking for. It allows to bypass
1376 routes defined between <b>host/router</b>. The principle is the same
1377 as route : <b>bypassRoute</b> contains list of links references of
1378 links that are in the path between src and dst.
1380 #### Attributes ####
1382 | Attribute name | Mandatory | Values | Description |
1383 | --------------- | --------- | ---------------------- | ----------- |
1384 | src | yes | String | The value given to the source AS's "id" attribute |
1385 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1386 | symmetrical | no | YES \| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1391 <bypassRoute src="host_1" dst="host_2">
1392 <link_ctn id="link_tmp"/>
1396 This example shows that link \c link_tmp (definition not displayed here) directly
1397 connects host \c host_1 to host \c host_2. Additionally, as the \c symmetrical
1398 attribute was not given, this route is presumed to be symmetrical.
1400 \subsection pb_baroex Basic Routing Example
1402 Let's say you have an AS named AS_Big that contains two other AS, AS_1
1403 and AS_2. If you want to make a host (h1) from AS_1 with another one
1404 (h2) from AS_2 then you'll have to proceed as follows:
1405 \li First, you have to ensure that a route is defined from h1 to the
1406 AS_1's exit gateway and from h2 to AS_2's exit gateway.
1407 \li Then, you'll have to define a route between AS_1 to AS_2. As those
1408 AS are both resources belonging to AS_Big, then it has to be done
1409 at AS_big level. To define such a route, you have to give the
1410 source AS (AS_1), the destination AS (AS_2), and their respective
1411 gateway (as the route is effectively defined between those two
1412 entry/exit points). Elements of this route can only be elements
1413 belonging to AS_Big, so links and routers in this route should be
1414 defined inside AS_Big. If you choose some shortest-path model,
1415 this route will be computed automatically.
1417 As said before, there are mainly 2 tags for routing :
1418 \li <b>ASroute</b>: to define routes between two <b>AS</b>
1419 \li <b>route</b>: to define routes between two <b>host/router</b>
1421 As we are dealing with routes between AS, it means that those we'll
1422 have some definition at AS_Big level. Let consider AS_1 contains 1
1423 host, 1 link and one router and AS_2 3 hosts, 4 links and one router.
1424 There will be a central router, and a cross-like topology. At the end
1425 of the crosses arms, you'll find the 3 hosts and the router that will
1426 act as a gateway. We have to define routes inside those two AS. Let
1427 say that AS_1 contains full routes, and AS_2 contains some Floyd
1428 routing (as we don't want to bother with defining all routes). As
1429 we're using some shortest path algorithms to route into AS_2, we'll
1430 then have to define some <b>route</b> to gives some topological
1431 information to SimGrid. Here is a file doing it all :
1434 <AS id="AS_Big" routing="Dijkstra">
1435 <AS id="AS_1" routing="Full">
1436 <host id="AS_1_host1" speed="1000000000"/>
1437 <link id="AS_1_link" bandwidth="1250000000" latency="5E-4"/>
1438 <router id="AS_1_gateway"/>
1439 <route src="AS_1_host1" dst="AS_1_gateway">
1440 <link_ctn id="AS_1_link"/>
1443 <AS id="AS_2" routing="Floyd">
1444 <host id="AS_2_host1" speed="1000000000"/>
1445 <host id="AS_2_host2" speed="1000000000"/>
1446 <host id="AS_2_host3" speed="1000000000"/>
1447 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1448 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1449 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1450 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1451 <router id="central_router"/>
1452 <router id="AS_2_gateway"/>
1453 <!-- routes providing topological information -->
1454 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1455 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1456 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1457 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1459 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1461 <ASroute src="AS_1" dst="AS_2"
1462 gw_src="AS_1_gateway"
1463 gw_dst="AS_2_gateway">
1464 <link_ctn id="backbone"/>
1469 \section pf_other_tags Tags not (directly) describing the platform
1471 The following tags can be used inside a \<platform\> tag even if they are not
1472 directly describing the platform:
1473 \li \ref pf_config "config": it allows you to pass some configuration stuff like, for
1474 example, the network model and so on. It follows the
1475 \li \ref pf_include "include": allows you to include another file into the current one.
1477 \subsection pf_config config
1479 The only purpose of this tag is to contain the \c prop tags, as described below.
1480 These tags will then configure the options as described by Section \ref options.
1483 #### Attributes ####
1485 | Attribute name | Mandatory | Values | Description |
1486 | --------------- | --------- | ---------------------- | ----------- |
1487 | id | yes | String | The identifier of the config tag when referring to id; this is basically useless, though. |
1489 #### Possible children ####
1491 Tag name | Description | Documentation
1492 ------------ | ----------- | -------------
1493 \<prop/\> | The prop tag allows you to define different configuration options following the attribute/value schema. See the \ref options page. | N/A
1498 <?xml version='1.0'?>
1499 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1500 <platform version="4">
1501 <config id="General">
1502 <prop id="maxmin/precision" value="0.000010"></prop>
1503 <prop id="cpu/optim" value="TI"></prop>
1504 <prop id="host/model" value="compound"></prop>
1505 <prop id="network/model" value="SMPI"></prop>
1506 <prop id="path" value="~/"></prop>
1507 <prop id="smpi/bw-factor" value="65472:0.940694;15424:0.697866;9376:0.58729"></prop>
1510 <AS id="AS0" routing="Full">
1514 \subsection pf_include include
1516 Even if it can be used in other contexts, this tag was originally created
1517 to be used with \ref pf_trace. The idea was to have a file describing the
1518 platform, and another file attaching traces of a given period to the platform.
1520 The drawback is that the file chuncks that will be included do not
1521 constitute valid XML files. This may explain why this feature was never really
1522 used in practice (as far as we know). Other mechanisms, such as the ability to load
1523 several platform files one after the other, could be considered in the future.
1525 In the meanwhile, the \c include tag allows you to import other platforms into your
1526 local file. This is done with the intention to help people
1527 combine their different AS and provide new platforms. Those files
1528 should contain XML that consists of
1529 \ref pf_include "include", \ref pf_cluster "cluster", \ref pf_peer "peer", \ref pf_As "AS", \ref pf_trace "trace", \ref pf_trace "tags".
1531 Do not forget to close the tag to make it work, or you will end up with an invalid XML file.
1533 #### Attributes ####
1535 | Attribute name | Mandatory | Values | Description |
1536 | --------------- | --------- | ---------------------- | ----------- |
1537 | file | yes | String | Filename of the path you want to include with either relative or absolute path. |
1542 The following example includes two files, clusterA.xml and clusterB.xml and
1543 combines them two one platform file; all hosts, routers etc. defined in
1544 each of them will then be usable.
1547 <?xml version='1.0'?>
1548 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid/simgrid.dtd">
1549 <platform version="4">
1550 <AS id="main" routing="Full">
1551 <include file="clusterA.xml"></include>
1552 <include file="clusterB.xml"></include>
1557 \subsection pf_trace trace and trace_connect
1559 Both tags are an alternate way to pass files containing information on
1560 availability, state etc. to an entity. (See also, for instance, Section \ref
1561 pf_host_churn "Churn", as described for the host entity.) Instead of referring
1562 to the file directly in the host, link, or cluster tag, you proceed by defining
1563 a trace with an id corresponding to a file, later a host/link/cluster, and
1564 finally using trace_connect you say that the file trace must be used by the
1571 <AS id="AS0" routing="Full">
1572 <host id="bob" speed="1000000000"/>
1574 <trace id="myTrace" file="bob.trace" periodicity="1.0"/>
1575 <trace_connect trace="myTrace" element="bob" kind="POWER"/>
1579 The order here is important. \c trace_connect must come
1580 after the elements \c trace and \c host, as both the host
1581 and the trace definition must be known when \c trace_connect
1582 is parsed; the order of \c trace and \c host is arbitrary.
1585 #### \c trace attributes ####
1588 | Attribute name | Mandatory | Values | Description |
1589 | --------------- | --------- | ---------------------- | ----------- |
1590 | id | yes | String | Identifier of this trace; this is the name you pass on to \c trace_connect. |
1591 | 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. |
1592 | trace_periodicity | yes | String | This is the same as for \ref pf_tag_host "hosts" (see there for details) |
1594 Here is an example of trace when no file name is provided:
1597 <trace id="myTrace" periodicity="1.0">
1604 #### \c trace_connect attributes ####
1606 | Attribute name | Mandatory | Values | Description |
1607 | --------------- | --------- | ---------------------- | ----------- |
1608 | kind | no | HOST_AVAIL\|POWER\|<br/>LINK_AVAIL\|BANDWIDTH\|LATENCY (Default: HOST_AVAIL) | Describes the kind of trace. |
1609 | trace | yes | String | Identifier of the referenced trace (specified of the trace's \c id attribute) |
1610 | element | yes | String | The identifier of the referenced entity as given by its \c id attribute |
1612 \section pf_hints Hints, tips and frequently requested features
1614 Now you should know at least the syntax and be able to create a
1615 platform by your own. However, after having ourselves wrote some platforms, there
1616 are some best practices you should pay attention to in order to
1617 produce good platform and some choices you can make in order to have
1618 faster simulations. Here's some hints and tips, then.
1620 @subsection Finding the platform example that you need
1622 Most platform files that we ship are in the @c examples/platforms
1623 folder. The good old @c grep tool can find the examples you need when
1624 wondering on a specific XML tag. Here is an example session searching
1625 for @ref pf_trace "trace_connect":
1628 % cd examples/platforms
1629 % grep -R -i -n --include="*.xml" "trace_connect" .
1630 ./two_hosts_platform_with_availability_included.xml:26:<trace_connect kind="SPEED" trace="A" element="Cpu A"/>
1631 ./two_hosts_platform_with_availability_included.xml:27:<trace_connect kind="HOST_AVAIL" trace="A_failure" element="Cpu A"/>
1632 ./two_hosts_platform_with_availability_included.xml:28:<trace_connect kind="SPEED" trace="B" element="Cpu B"/>
1633 ./two_hosts.xml:17: <trace_connect trace="Tremblay_power" element="Tremblay" kind="SPEED"/>
1636 \subsection pf_as_h AS Hierarchy
1637 The AS design allows SimGrid to go fast, because computing route is
1638 done only for the set of resources defined in this AS. If you're using
1639 only a big AS containing all resource with no AS into it and you're
1640 using Full model, then ... you'll loose all interest into it. On the
1641 other hand, designing a binary tree of AS with, at the lower level,
1642 only one host, then you'll also loose all the good AS hierarchy can
1643 give you. Remind you should always be "reasonable" in your platform
1644 definition when choosing the hierarchy. A good choice if you try to
1645 describe a real life platform is to follow the AS described in
1646 reality, since this kind of trade-off works well for real life
1649 \subsection pf_exit_as Exit AS: why and how
1650 Users that have looked at some of our platforms may have notice a
1651 non-intuitive schema ... Something like that :
1655 <AS id="AS_4" routing="Full">
1656 <AS id="exitAS_4" routing="Full">
1657 <router id="router_4"/>
1659 <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"/>
1660 <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"/>
1661 <link id="4_1" bandwidth="2250000000" latency="5E-5"/>
1662 <link id="4_2" bandwidth="2250000000" latency="5E-5"/>
1663 <link id="bb_4" bandwidth="2250000000" latency="5E-4"/>
1664 <ASroute src="cl_4_1"
1666 gw_src="c_4_1-cl_4_1_router"
1667 gw_dst="c_4_2-cl_4_2_router">
1668 <link_ctn id="4_1"/>
1669 <link_ctn id="bb_4"/>
1670 <link_ctn id="4_2"/>
1672 <ASroute src="cl_4_1"
1674 gw_src="c_4_1-cl_4_1_router"
1676 <link_ctn id="4_1"/>
1677 <link_ctn id="bb_4"/>
1679 <ASroute src="cl_4_2"
1681 gw_src="c_4_2-cl_4_2_router"
1683 <link_ctn id="4_2"/>
1684 <link_ctn id="bb_4"/>
1689 In the AS_4, you have an exitAS_4 defined, containing only one router,
1690 and routes defined to that AS from all other AS (as cluster is only a
1691 shortcut for an AS, see cluster description for details). If there was
1692 an upper AS, it would define routes to and from AS_4 with the gateway
1693 router_4. It's just because, as we did not allowed (for performances
1694 issues) to have routes from an AS to a single host/router, you have to
1695 enclose your gateway, when you have AS included in your AS, within an
1696 AS to define routes to it.
1698 \subsection pf_P2P_tags P2P or how to use coordinates
1699 SimGrid allows you to use some coordinated-based system, like vivaldi,
1700 to describe a platform. The main concept is that you have some peers
1701 that are located somewhere: this is the function of the
1702 <b>coordinates</b> of the \<peer\> or \<host\> tag. There's nothing
1703 complicated in using it, here is an example:
1706 <?xml version='1.0'?>
1707 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1708 <platform version="4">
1710 <AS id="AS0" routing="Vivaldi">
1711 <host id="100030591" coordinates="25.5 9.4 1.4" speed="1.5Gf" />
1712 <host id="100036570" coordinates="-12.7 -9.9 2.1" speed="7.3Gf" />
1714 <host id="100429957" coordinates="17.5 6.7 18.8" speed="8.3Gf" />
1719 Coordinates are then used to calculate latency (in microseconds)
1720 between two hosts by calculating the distance between the two hosts
1721 coordinates with the following formula: distance( (x1, y1, z1), (x2,
1722 y2, z2) ) = euclidian( (x1,y1), (x2,y2) ) + abs(z1) + abs(z2)
1724 In other words, we take the euclidian distance on the two first
1725 dimensions, and then add the absolute values found on the third
1726 dimension. This may seem strange, but it was found to allow better
1727 approximations of the latency matrices (see the paper describing
1730 Note that the previous example defines a routing directly between hosts but it could be also used to define a routing between AS.
1731 That is for example what is commonly done when using peers (see Section \ref pf_peer).
1733 <?xml version='1.0'?>
1734 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1735 <platform version="4">
1737 <AS id="AS0" routing="Vivaldi">
1738 <peer id="peer-0" coordinates="173.0 96.8 0.1" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us"/>
1739 <peer id="peer-1" coordinates="247.0 57.3 0.6" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1740 <peer id="peer-2" coordinates="243.4 58.8 1.4" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1744 In such a case though, we connect the AS created by the <b>peer</b> tag with the Vivaldi routing mechanism.
1745 This means that to route between AS1 and AS2, it will use the coordinates of router_AS1 and router_AS2.
1746 This is currently a convention and we may offer to change this convention in the DTD later if needed.
1747 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.
1750 \subsection pf_routing_howto_choose_wisely Choosing wisely the routing model to use
1753 Choosing wisely the routing model to use can significantly fasten your
1754 simulation/save your time when writing the platform/save tremendous
1755 disk space. Here is the list of available model and their
1756 characteristics (lookup : time to resolve a route):
1758 \li <b>Full</b>: Full routing data (fast, large memory requirements,
1760 \li <b>Floyd</b>: Floyd routing data (slow initialization, fast
1761 lookup, lesser memory requirements, shortest path routing only).
1762 Calculates all routes at once at the beginning.
1763 \li <b>Dijkstra</b>: Dijkstra routing data (fast initialization, slow
1764 lookup, small memory requirements, shortest path routing only).
1765 Calculates a route when necessary.
1766 \li <b>DijkstraCache</b>: Dijkstra routing data (fast initialization,
1767 fast lookup, small memory requirements, shortest path routing
1768 only). Same as Dijkstra, except it handles a cache for latest used
1770 \li <b>None</b>: No routing (usable with Constant network only).
1771 Defines that there is no routes, so if you try to determine a
1772 route without constant network within this AS, SimGrid will raise
1774 \li <b>Vivaldi</b>: Vivaldi routing, so when you want to use coordinates
1775 \li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
1778 \subsection pf_switch I want to describe a switch but there is no switch tag!
1780 Actually we did not include switch tag. But when you're trying to
1781 simulate a switch, assuming
1782 fluid bandwidth models are used (which SimGrid uses by default unless
1783 ns-3 or constant network models are activated), the limiting factor is
1784 switch backplane bandwidth. So, essentially, at least from
1785 the simulation perspective, a switch is similar to a
1786 link: some device that is traversed by flows and with some latency and
1787 so,e maximum bandwidth. Thus, you can simply simulate a switch as a
1789 can be connected to this "switch", which is then included in routes just
1793 \subsection pf_multicabinets I want to describe multi-cabinets clusters!
1795 You have several possibilities, as usual when modeling things. If your
1796 cabinets are homogeneous and the intercabinet network negligible for
1797 your study, you should just create a larger cluster with all hosts at
1800 In the rare case where your hosts are not homogeneous between the
1801 cabinets, you can create your cluster completely manually. For that,
1802 create an As using the Cluster routing, and then use one
1803 <cabinet> for each cabinet. This cabinet tag can only be used an
1804 As using the Cluster routing schema, and creating
1806 Be warned that creating a cluster manually from the XML with
1807 <cabinet>, <backbone> and friends is rather tedious. The
1808 easiest way to retrieve some control of your model without diving into
1809 the <cluster> internals is certainly to create one separate
1810 <cluster> per cabinet and interconnect them together. This is
1811 what we did in the G5K example platform for the Graphen cluster.
1813 \subsection pf_platform_multipath I want to express multipath routing in platform files!
1815 It is unfortunately impossible to express the fact that there is more
1816 than one routing path between two given hosts. Let's consider the
1817 following platform file:
1820 <route src="A" dst="B">
1823 <route src="B" dst="C">
1826 <route src="A" dst="C">
1831 Although it is perfectly valid, it does not mean that data traveling
1832 from A to C can either go directly (using link 3) or through B (using
1833 links 1 and 2). It simply means that the routing on the graph is not
1834 trivial, and that data do not following the shortest path in number of
1835 hops on this graph. Another way to say it is that there is no implicit
1836 in these routing descriptions. The system will only use the routes you
1837 declare (such as <route src="A" dst="C"><link_ctn
1838 id="3"/></route>), without trying to build new routes by aggregating
1841 You are also free to declare platform where the routing is not
1842 symmetrical. For example, add the following to the previous file:
1845 <route src="C" dst="A">
1851 This makes sure that data from C to A go through B where data from A
1852 to C go directly. Don't worry about realism of such settings since
1853 we've seen ways more weird situation in real settings (in fact, that's
1854 the realism of very regular platforms which is questionable, but
1855 that's another story).
