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_res Resource description
109 \subsection pf_res_computing Computing Resources
111 \subsubsection pf_tag_host <host>
113 An host is the computing resource on which an actor can execute.
115 Attribute | Values | Description
116 ----------------- | -------------------------------------- | -----------
117 id | String (mandatory) | The identifier of the host. facilitates referring to this AS.
118 speed | double (mandatory) | Computational power of every core of this host in FLOPS (must be positive)
119 core | int (defaults to 1) | Number of cores (see @ref howto_multicore)
120 state | optionally "OFF" | If set to OFF, the host is initially turned off.
121 availability_file | File name (optional) | (Relative or absolute) filename to use as input; must contain availability traces for this host. The syntax of this file is defined below.
122 state_file | File name (optional) | Same mechanism as availability_file.<br />
123 coordinates | String (mandatory when using Vivaldi routing) | The coordinates of this host (see @ref pf_P2P_tags).
124 pstate | Double (Defaults to 0) | FIXME: Not yet documented.
126 #### Included tags ####
128 - @ref pf_tag_mount Specifies the storages mounted on that host
129 - @ref pf_tag_prop Specifies a user-defined property of that host, that you can retrieve with MSG_host_get_property_value() or simgrid::s4u::Host::property().
134 <host id="host1" speed="1000000000"/>
135 <host id="host2" speed="1000000000">
136 <prop id="color" value="blue"/>
137 <prop id="rendershape" value="square"/>
141 \anchor pf_host_dynamism
142 ### Expressing dynamism ###
144 SimGrid provides mechanisms to change a hosts' availability over
145 time, using the ``availability_file`` attribute to the ``\<host\>`` tag
146 and a separate text file whose syntax is exemplified below.
148 #### Adding a trace file ####
151 <platform version="4">
152 <host id="bob" speed="500Gf" availability_file="bob.trace" />
156 #### Example of "bob.trace" file ####
165 Let us begin to explain this example by looking at line 2. (Line 1 will become clear soon).
166 The first column describes points in time, in this case, time 0. The second column
167 describes the relative amount of power this host is able to deliver (relative
168 to the maximum performance specified in the ``\<host\>`` tag). (Clearly, the
169 second column needs to contain values that are not smaller than 0 and not larger than 1).
170 In this example, our host will deliver 500 Mflop/s at time 0, as 500 Mflop/s is the
171 maximum performance of this host. At time 11.0, it will
172 deliver half of its maximum performance, i.e., 250 Mflop/s until time 20.0 when it will
173 will start delivering 80\% of its power. In this example, this amounts to 400 Mflop/s.
175 Since the periodicity in line 1 was set to be 1.0, i.e., 1 timestep, this host will
176 continue to provide 500 Mflop/s from time 21. From time 32 it will provide 250 MFlop/s and so on.
178 \anchor pf_host_churn
179 ### Expressing churn ###
181 To express the fact that a host can change state over time (as in P2P
182 systems, for instance), it is possible to use a file describing the time
183 at which the host is turned on or off. An example of the content
184 of such a file is presented below.
186 #### Adding a state file ####
189 <platform version="4">
190 <host id="bob" power="500Gf" state_file="bob.fail" />
194 #### Example of "bob.fail" file ####
202 A zero value means <b>down</b> (i.e., OFF) while a positive one means <b>up and
203 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
204 turned off and at time 2.0, it is turned on again until time 12 (2 plus the
205 periodicity 10). It will be turned off again at time 13.0 until time 23.0, and
209 \subsubsection pf_tag_cluster <cluster>
211 ``<cluster />`` represents a machine-cluster. It is most commonly used
212 when one wants to define many hosts and a network quickly. Technically,
213 ``cluster`` is a meta-tag: <b>from the inner SimGrid point of
214 view, a cluster is an AS where some optimized routing is defined</b>.
215 The default inner organization of the cluster is as follow:
221 ____________|__________|_____________ backbone
223 l0| l1| l2| l97| l96 | | l99
229 Here, a set of <b>host</b>s is defined. Each of them has a <b>link</b>
230 to a central backbone (backbone is a link itself, as a link can
231 be used to represent a switch, see the switch / link section
232 below for more details about it). A <b>router</b> allows to connect a
233 <b>cluster</b> to the outside world. Internally,
234 SimGrid treats a cluster as an AS containing all hosts: the router is the default
235 gateway for the cluster.
237 There is an alternative organization, which is as follows:
251 The principle is the same, except that there is no backbone. This representation
252 can be obtained easily: just do not set the bb_* attributes.
255 Attribute name | Mandatory | Values | Description
256 --------------- | --------- | ------ | -----------
257 id | yes | string | The identifier of the cluster. Facilitates referring to this cluster.
258 prefix | yes | string | Each node of the cluster has to have a name. This name will be prefixed with this prefix.
259 suffix | yes | string | Each node of the cluster will be suffixed with this suffix
260 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.
261 speed | yes | int | Same as the ``speed`` attribute of the ``\<host\>`` tag.
262 core | no | int (default: 1) | Same as the ``core`` attribute of the ``\<host\>`` tag.
263 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
264 lat | yes | int | Latency for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
265 sharing_policy | no | string | Sharing policy for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
266 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>).
267 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>).
268 bb_sharing_policy | no | string | Sharing policy for the backbone (if any). See <b>link</b> section for syntax/details.
269 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>
270 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".
271 loopback_lat | no | int | Latency for loopback (if any). See <b>link</b> section for syntax/details. See loopback_bw for more info.
272 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.
273 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".
276 the router name is defined as the resulting String in the following
280 router_name = prefix + clusterId + "_router" + suffix;
284 #### Cluster example ####
286 Consider the following two (and independent) uses of the ``cluster`` tag:
289 <cluster id="my_cluster_1" prefix="" suffix="" radical="0-262144"
290 speed="1e9" bw="125e6" lat="5E-5"/>
292 <cluster id="my_cluster_2" prefix="c-" suffix=".me" radical="0-99"
293 speed="1e9" bw="125e6" lat="5E-5"
294 bb_bw="2.25e9" bb_lat="5E-4"/>
297 The second example creates one router and 100 machines with the following names:
299 c-my_cluster_2_router.me
307 \subsubsection pf_cabinet <cabinet>
310 This tag is only available when the routing mode of the AS
311 is set to ``Cluster``.
313 The ``<cabinet />`` tag is, like the \ref pf_tag_cluster "<cluster>" tag,
314 a meta-tag. This means that it is simply a shortcut for creating a set of (homogenous) hosts and links quickly;
315 unsurprisingly, this tag was introduced to setup cabinets in data centers quickly. Unlike
316 <cluster>, however, the <cabinet> assumes that you create the backbone
317 and routers yourself; see our examples below.
321 Attribute name | Mandatory | Values | Description
322 --------------- | --------- | ------ | -----------
323 id | yes | string | The identifier of the cabinet. Facilitates referring to this cluster.
324 prefix | yes | string | Each node of the cabinet has to have a name. This name will be prefixed with this prefix.
325 suffix | yes | string | Each node of the cabinet will be suffixed with this suffix
326 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.
327 speed | yes | int | Same as the ``speed`` attribute of the \ref pf_tag_host "<host>" tag.
328 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
329 lat | yes | int | Latency for the links between nodes and backbone (if any). See the \ref pf_tag_link "link section" for syntax/details.
332 Please note that as of now, it is impossible to change attributes such as,
333 amount of cores (always set to 1), the initial state of hosts/links
334 (always set to ON), the sharing policy of the links (always set to \ref pf_sharing_policy_fullduplex "FULLDUPLEX").
338 The following example was taken from ``examples/platforms/meta_cluster.xml`` and
339 shows how to use the cabinet tag.
342 <AS id="my_cluster1" routing="Cluster">
343 <cabinet id="cabinet1" prefix="host-" suffix=".cluster1"
344 speed="1Gf" bw="125MBps" lat="100us" radical="1-10"/>
345 <cabinet id="cabinet2" prefix="host-" suffix=".cluster1"
346 speed="1Gf" bw="125MBps" lat="100us" radical="11-20"/>
347 <cabinet id="cabinet3" prefix="host-" suffix=".cluster1"
348 speed="1Gf" bw="125MBps" lat="100us" radical="21-30"/>
350 <backbone id="backbone1" bandwidth="2.25GBps" latency="500us"/>
355 Please note that you must specify the \ref pf_backbone "<backbone>"
356 tag by yourself; this is not done automatically and there are no checks
357 that ensure this backbone was defined.
359 The hosts generated in the above example are named host-1.cluster, host-2.cluster1
363 \subsubsection pf_peer \<peer\> (Vivaldi netzones only)
365 This tag represents a peer, as in Peer-to-Peer (P2P) networks. This
366 can only be used in Vivaldi NetZones. It creates the following
367 resources to the NetZone:
370 \li Two links: One for download and one for upload. This is
371 convenient to use and simulate stuff under the last mile model (e.g., ADSL peers).
372 \li It connects the two links to the host
376 Attribute name | Mandatory | Values | Description
377 --------------- | --------- | ------ | -----------
378 id | yes | string | The identifier of the peer. Facilitates referring to this peer.
379 speed | yes | int | See the description of the ``host`` tag for this attribute
380 bw_in | yes | int | Bandwidth of the private downstream link
381 bw_out | yes | int | Bandwidth of the private upstream link
382 coordinates | no | string | Coordinates of the gateway for this peer. Example value: 12.8 14.4 6.4
383 sharing_policy | no | SHARED\|FULLDUPLEX (default: FULLDUPLEX) | Sharing policy for links. See <b>link</b> description for details.
384 availability_file| no | string | Availability file for the peer. Same as host availability file. See <b>host</b> description for details.
385 state_file | no | string | State file for the peer. Same as host state file. See <b>host</b> description for details.
388 The communication latency between an host A=(xA,yA,zA) and an host
389 B=(xB,yB,zB) is computed as follows:
391 latency = sqrt( (xA-xB)² + (yA-yB)² ) + zA + zB
393 See the documentation of simgrid::kernel::routing::VivaldiZone for
394 details on how the latency is computed from the coordinate, and on the
395 the up and down bandwidth are used.
397 \subsection pf_ne Network equipments
399 There are two tags at all times available to represent network entities and
400 several other tags that are available only in certain contexts.
401 1. ``<link>``: Represents a entity that has a limited bandwidth, a
402 latency, and that can be shared according to TCP way to share this
405 The concept of links in SimGrid may not be intuitive, as links are not
406 limited to connecting (exactly) two entities; in fact, you can have more than
407 two equipments connected to it. (In graph theoretical terms: A link in
408 SimGrid is not an edge, but a hyperedge)
410 2. ``<router/>``: Represents an entity that a message can be routed
411 to, but that is unable to execute any code. In SimGrid, routers have also
412 no impact on the performance: Routers do not limit any bandwidth nor
413 do they increase latency. As a matter of fact, routers are (almost) ignored
414 by the simulator when the simulation has begun.
416 3. ``<backbone/>``: This tag is only available when the containing AS is
417 used as a cluster (i.e., mode="Cluster")
420 If you want to represent an entity like a switch, you must use ``<link>`` (see section). Routers are used
421 to run some routing algorithm and determine routes (see Section \ref pf_routing for details).
423 \subsubsection pf_router <router/>
425 As said before, <b>router</b> is used only to give some information
426 for routing algorithms. So, it does not have any attributes except :
430 Attribute name | Mandatory | Values | Description
431 --------------- | --------- | ------ | -----------
432 id | yes | string | The identifier of the router to be used when referring to it.
433 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.
438 <router id="gw_dc1_horizdist"/>
441 \subsubsection pf_tag_link <link>
443 Network links can represent one-hop network connections. They are
444 characterized by their id and their bandwidth; links can (but may not) be subject
449 Attribute name | Mandatory | Values | Description
450 --------------- | --------- | ------ | -----------
451 id | yes | string | The identifier of the link to be used when referring to it.
452 bandwidth | yes | int | Maximum bandwidth for this link, given in bytes/s
453 latency | no | double (default: 0.0) | Latency for this link.
454 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.
455 state | no | ON\|OFF (default: ON) | Allows you to to turn this link on or off (working / not working)
456 bandwidth_file | no | string | Allows you to use a file as input for bandwidth.
457 latency_file | no | string | Allows you to use a file as input for latency.
458 state_file | no | string | Allows you to use a file as input for states.
461 #### Possible shortcuts for ``latency`` ####
463 When using the latency attribute, you can specify the latency by using the scientific
464 notation or by using common abbreviations. For instance, the following three tags
468 <link id="LINK1" bandwidth="125000000" latency="5E-6"/>
469 <link id="LINK1" bandwidth="125000000" latency="5us"/>
470 <link id="LINK1" bandwidth="125000000" latency="0.000005"/>
473 Here, the second tag uses "us", meaning "microseconds". Other shortcuts are:
475 Name | Abbreviation | Time (in seconds)
476 ---- | ------------ | -----------------
477 Week | w | 7 * 24 * 60 * 60
478 Day | d | 24 * 60 * 60
482 Millisecond | ms | 0.001 = 10^(-3)
483 Microsecond | us | 0.000001 = 10^(-6)
484 Nanosecond | ns | 0.000000001 = 10^(-9)
485 Picosecond | ps | 0.000000000001 = 10^(-12)
487 #### Sharing policy ####
489 \anchor sharing_policy_shared
490 By default a network link is \b SHARED, i.e., if two or more data flows go
491 through a link, the bandwidth is shared fairly among all data flows. This
492 is similar to the sharing policy TCP uses.
494 \anchor pf_sharing_policy_fatpipe
495 On the other hand, if a link is defined as a \b FATPIPE,
496 each flow going through this link will be provided with the complete bandwidth,
497 i.e., no sharing occurs and the bandwidth is only limiting each flow individually.
498 Please note that this is really on a per-flow basis, not only on a per-host basis!
499 The complete bandwidth provided by this link in this mode
500 is ``number_of_flows*bandwidth``, with at most ``bandwidth`` being available per flow.
502 Using the FATPIPE mode allows to model backbones that won't affect performance
505 \anchor pf_sharing_policy_fullduplex
506 The last mode available is \b FULLDUPLEX. This means that SimGrid will
507 automatically generate two links (one carrying the suffix _UP and the other the
508 suffix _DOWN) for each ``<link>`` tag. This models situations when the direction
509 of traffic is important.
512 Transfers from one side to the other will interact similarly as
513 TCP when ACK returning packets circulate on the other direction. More
514 discussion about it is available in the description of link_ctn description.
516 In other words: The SHARED policy defines a physical limit for the bandwidth.
517 The FATPIPE mode defines a limit for each application,
518 with no upper total limit.
521 Tip: By using the FATPIPE mode, you can model big backbones that
522 won't affect performance (except latency).
527 <link id="SWITCH" bandwidth="125000000" latency="5E-5" sharing_policy="FATPIPE" />
530 #### Expressing dynamism and failures ####
532 Similar to hosts, it is possible to declare links whose state, bandwidth
533 or latency changes over time (see Section \ref pf_host_dynamism for details).
535 In the case of network links, the ``bandwidth`` and ``latency`` attributes are
536 replaced by the ``bandwidth_file`` and ``latency_file`` attributes.
537 The following XML snippet demonstrates how to use this feature in the platform
538 file. The structure of the files "link1.bw" and "link1.lat" is shown below.
541 <link id="LINK1" state_file="link1.fail" bandwidth="80000000" latency=".0001" bandwidth_file="link1.bw" latency_file="link1.lat" />
545 Even if the syntax is the same, the semantic of bandwidth and latency
546 trace files differs from that of host availability files. For bandwidth and
547 latency, the corresponding files do not
548 express availability as a fraction of the available capacity but directly in
549 bytes per seconds for the bandwidth and in seconds for the latency. This is
550 because most tools allowing to capture traces on real platforms (such as NWS)
551 express their results this way.
553 ##### Example of "link1.bw" file #####
561 In this example, the bandwidth changes repeatedly, with all changes
562 being repeated every 12 seconds.
564 At the beginning of the the simulation, the link's bandwidth is 80,000,000
565 B/s (i.e., 80 Mb/s); this value was defined in the XML snippet above.
566 After four seconds, it drops to 40 Mb/s (line 2), and climbs
567 back to 60 Mb/s after another 4 seconds (line 3). The value does not change any
568 more until the end of the period, that is, after 12 seconds have been simulated).
569 At this point, periodicity kicks in and this behavior is repeated: Seconds
570 12-16 will experience 80 Mb/s, 16-20 40 Mb/s etc.).
572 ##### Example of "link1.lat" file #####
581 In this example, the latency varies with a period of 5 seconds.
582 In the xml snippet above, the latency is initialized to be 0.0001s (100µs). This
583 value will be kept during the first second, since the latency_file contains
584 changes to this value at second one, two and three.
585 At second one, the value will be 0.001, i.e., 1ms. One second later it will
586 be adjusted to 0.01 (or 10ms) and one second later it will be set again to 1ms. The
587 value will not change until second 5, when the periodicity defined in line 1
588 kicks in. It then loops back, starting at 100µs (the initial value) for one second.
590 #### The ``<prop/>`` tag ####
592 Similar to the ``<host>`` tag, a link may also contain the ``<prop/>`` tag; see the host
593 documentation (Section \ref pf_host) for an example.
596 \subsubsection pf_backbone <backbone/>
599 This tag is <b>only available</b> when the containing AS uses the "Cluster" routing mode!
601 Using this tag, you can designate an already existing link to be a backbone.
603 Attribute name | Mandatory | Values | Description
604 --------------- | --------- | ------ | -----------
605 id | yes | string | Name of the link that is supposed to act as a backbone.
607 \subsection pf_storage Storage
610 This is a prototype version that should evolve quickly, hence this
611 is just some doc valuable only at the time of writing.
612 This section describes the storage management under SimGrid ; nowadays
613 it's only usable with MSG. It relies basically on linux-like concepts.
614 You also may want to have a look to its corresponding section in
615 @ref msg_file ; access functions are organized as a POSIX-like
618 \subsubsection pf_sto_conc Storage - Main Concepts
620 The storage facilities implemented in SimGrid help to model (and account for)
621 storage devices, such as tapes, hard-drives, CD or DVD devices etc.
622 A typical situation is depicted in the figure below:
624 \image html ./webcruft/storage_sample_scenario.png
625 \image latex ./webcruft/storage_sample_scenario.png "storage_sample_scenario" width=\textwidth
627 In this figure, two hosts called Bob and Alice are interconnected via a network
628 and each host is physically attached to a disk; it is not only possible for each host to
629 mount the disk they are attached to directly, but they can also mount disks
630 that are in a remote location. In this example, Bob mounts Alice's disk remotely
631 and accesses the storage via the network.
633 SimGrid provides 3 different entities that can be used to model setups
634 that include storage facilities:
636 Entity name | Description
637 --------------- | -----------
638 \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).
639 \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.
640 \ref pf_tag_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).
643 \anchor pf_storage_content_file
644 ### %Storage Content File ###
646 In order to assess exactly how much time is spent reading from the storage,
647 SimGrid needs to know what is stored on the storage device (identified by distinct (file-)name, like in a file system)
648 and what size this content has.
651 The content file is never changed by the simulation; it is parsed once
652 per simulation and kept in memory afterwards. When the content of the
653 storage changes, only the internal SimGrid data structures change.
655 \anchor pf_storage_content_file_structure
656 #### Structure of a %Storage Content File ####
658 Here is an excerpt from two storage content file; if you want to see the whole file, check
659 the file ``examples/platforms/content/storage_content.txt`` that comes with the
662 SimGrid essentially supports two different formats: UNIX-style filepaths should
663 follow the well known format:
666 /lib/libsimgrid.so.3.6.2 12710497
670 /bin/simgrid_update_xml 5018
671 /bin/graphicator 66986
672 /bin/simgrid-colorizer 2993
677 Windows filepaths, unsurprisingly, use the windows style:
680 \Windows\avastSS.scr 41664
681 \Windows\bfsvc.exe 75264
682 \Windows\bootstat.dat 67584
683 \Windows\CoreSingleLanguage.xml 31497
685 \Windows\dchcfg64.exe 335464
686 \Windows\dcmdev64.exe 93288
690 The different file formats come at a cost; in version 3.12 (and most likely
691 in later versions, too), copying files from windows-style storages to unix-style
692 storages (and vice versa) is not supported.
694 \anchor pf_storage_content_file_create
695 #### Generate a %Storage Content File ####
697 If you want to generate a storage content file based on your own filesystem (or at least a filesystem you have access to),
698 try running this command (works only on unix systems):
701 find . -type f -exec ls -1s --block=1 {} \; 2>/dev/null | awk '{ print $2 " " $1}' > ./content.txt
704 \subsubsection pf_storage_entities The Storage Entities
706 These are the entities that you can use in your platform files to include
707 storage in your model. See also the list of our \ref pf_storage_example_files "example files";
708 these might also help you to get started.
710 \anchor pf_storage_entity_storage_type
711 #### \<storage_type\> ####
713 Attribute name | Mandatory | Values | Description
714 --------------- | --------- | ------ | -----------
715 id | yes | string | Identifier of this storage_type; used when referring to it
716 model | yes | string | For reasons of future backwards compatibility only; specifies the name of the model for the storage that should be used
717 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)
718 content | yes | string | Path to a \ref pf_storage_content_file "Storage Content File" on your system. This file must exist.
719 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"!)
721 This tag must contain some predefined model properties, specified via the <model_prop> tag. Here is a list,
722 see below for an example:
724 Property id | Mandatory | Values | Description
725 --------------- | --------- | ------ | -----------
726 Bwrite | yes | string | Bandwidth for write access; in B/s (but you can also specify e.g. "30MBps")
727 Bread | yes | string | Bandwidth for read access; in B/s (but you can also specify e.g. "30MBps")
728 Bconnexion | yes | string | Throughput (of the storage connector) in B/s.
731 A storage_type can also contain the <b><prop></b> tag. The <prop> tag allows you
732 to associate additional information to this <storage_type> and follows the
733 attribute/value schema; see the example below. You may want to use it to give information to
734 the tool you use for rendering your simulation, for example.
736 Here is a complete example for the ``storage_type`` tag:
738 <storage_type id="single_HDD" model="linear_no_lat" size="4000" content_type="txt_unix">
739 <model_prop id="Bwrite" value="30MBps" />
740 <model_prop id="Bread" value="100MBps" />
741 <model_prop id="Bconnection" value="150MBps" />
742 <prop id="Brand" value="Western Digital" />
746 @subsubsection pf_tag_storage <storage>
748 Attributes | Mandatory | Values | Description
749 -------------- | --------- | ------ | -----------
750 id | yes | string | Identifier of this ``storage``; used when referring to it
751 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.
752 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)
753 content | no | string | When specified, overwrites the content attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
754 content_type | no | string | When specified, overwrites the content_type attribute of \ref pf_storage_entity_storage_type "\<storage_type\>"
756 Here are two examples:
759 <storage id="Disk1" typeId="single_HDD" attach="bob" />
761 <storage id="Disk2" typeId="single_SSD"
762 content="content/win_storage_content.txt"
763 content_type="txt_windows" attach="alice" />
766 The first example is straightforward: A disk is defined and called "Disk1"; it is
767 of type "single_HDD" (shown as an example of \ref pf_storage_entity_storage_type "\<storage_type\>" above) and attached
768 to a host called "bob" (the definition of this host is omitted here).
770 The second storage is called "Disk2", is still of the same type as Disk1 but
771 now specifies a new content file (so the contents will be different from Disk1)
772 and the filesystem uses the windows style; finally, it is attached to a second host,
773 called alice (which is again not defined here).
775 \subsubsection pf_tag_mount <mount>
777 | Attribute | Mandatory | Values | Description |
778 | ----------- | ----------- | -------- | ------------- |
779 | id | yes | string | Refers to a \ref pf_storage_entity_storage "<storage>" entity that will be mounted on that computer |
780 | name | yes | string | Path/location to/of the logical reference (mount point) of this disk
782 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)
783 is; this location is specified by the ``name`` attribute.
785 Here is a simple example, taken from the file ``examples/platform/storage.xml``:
788 <storage_type id="single_SSD" model="linear_no_lat" size="500GiB">
789 <model_prop id="Bwrite" value="60MBps" />
790 <model_prop id="Bread" value="200MBps" />
791 <model_prop id="Bconnection" value="220MBps" />
794 <storage id="Disk2" typeId="single_SSD"
795 content="content/win_storage_content.txt"
796 content_type="txt_windows" attach="alice" />
797 <storage id="Disk4" typeId="single_SSD"
798 content="content/small_content.txt"
799 content_type="txt_unix" attach="denise"/>
801 <host id="alice" speed="1Gf">
802 <mount storageId="Disk2" name="c:"/>
805 <host id="denise" speed="1Gf">
806 <mount storageId="Disk2" name="c:"/>
807 <mount storageId="Disk4" name="/home"/>
811 This example is quite interesting, as the same device, called "Disk2", is mounted by
812 two hosts at the same time! Note, however, that the host called ``alice`` is actually
813 attached to this storage, as can be seen in the \ref pf_storage_entity_storage "<storage>"
814 tag. This means that ``denise`` must access this storage through the network, but SimGrid automatically takes
815 care of that for you.
817 Furthermore, this example shows that ``denise`` has mounted two storages with different
818 filesystem types (unix and windows). In general, a host can mount as many storage devices as
822 Again, the difference between ``attach`` and ``mount`` is simply that
823 an attached storage is always physically inside (or connected to) that machine;
824 for instance, a USB stick is attached to one and only one machine (where it's plugged-in)
825 but it can only be mounted on others, as mounted storage can also be a remote location.
827 ###### Example files #####
829 \verbinclude example_filelist_xmltag_mount
831 \subsubsection pf_storage_example_files Example files
833 Several examples were already discussed above; if you're interested in full examples,
834 check the the following platforms:
836 1. ``examples/platforms/storage.xml``
837 2. ``examples/platforms/remote_io.xml``
839 If you're looking for some examplary C code, you may find the source code
840 available in the directory ``examples/msg/io/`` useful.
842 \subsubsection pf_storage_examples_modelling Modelling different situations
844 The storage functionality of SimGrid is type-agnostic, that is, the implementation
845 does not presume any type of storage, such as HDDs/SSDs, RAM,
846 CD/DVD devices, USB sticks etc.
848 This allows the user to apply the simulator for a wide variety of scenarios; one
849 common scenario would be the access of remote RAM.
851 #### Modelling the access of remote RAM ####
853 How can this be achieved in SimGrid? Let's assume we have a setup where three hosts
854 (HostA, HostB, HostC) need to access remote RAM:
864 An easy way to model this scenario is to setup and define the RAM via the
865 \ref pf_storage_entity_storage "storage" and \ref pf_storage_entity_storage_type "storage type"
866 entities and attach it to a remote dummy host; then, every host can have their own links
867 to this host (modelling for instance certain scenarios, such as PCIe ...)
872 RAM - Dummy -- Host B
877 Now, if read from this storage, the host that mounts this storage
878 communicates to the dummy host which reads from RAM and
879 sends the information back.
882 \section pf_routing Routing
884 To achieve high performance, the routing tables used within SimGrid are
885 static. This means that routing between two nodes is calculated once
886 and will not change during execution. The SimGrid team chose to use this
887 approach as it is rare to have a real deficiency of a resource;
888 most of the time, a communication fails because the links experience too much
889 congestion and hence, your connection stops before the timeout or
890 because the computer designated to be the destination of that message
893 We also chose to use shortest paths algorithms in order to emulate
894 routing. Doing so is consistent with the reality: [RIP](https://en.wikipedia.org/wiki/Routing_Information_Protocol),
895 [OSPF](https://en.wikipedia.org/wiki/Open_Shortest_Path_First), [BGP](https://en.wikipedia.org/wiki/Border_Gateway_Protocol)
896 are all calculating shortest paths. They do require some time to converge, but
897 eventually, when the routing tables have stabilized, your packets will follow
900 \subsection pf_tag_zone <zone>
902 Before SimGrid v3.16, networking zones used to be called Autonomous
903 Systems, but this was misleading as zones may include other zones in a
904 hierarchical manner. If you find any remaining reference to ASes,
905 please report this as a bug.
907 Attribute | Value | Description
908 ----------- | ------------------------------------------------- | ----------------------------------------------
909 id | String (mandatory) | The identifier of this zone (must be unique)
910 routing | One of the existing routing algorithm (mandatory) | See Section \ref pf_rm for details.
914 <AS id="AS0" routing="Full">
915 <host id="host1" speed="1000000000"/>
916 <host id="host2" speed="1000000000"/>
917 <link id="link1" bandwidth="125000000" latency="0.000100"/>
918 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
922 In this example, AS0 contains two hosts (host1 and host2). The route
923 between the hosts goes through link1.
925 \subsection pf_rm Routing models
927 For each AS, you must define explicitly which routing model will
928 be used. There are 3 different categories for routing models:
930 1. \ref pf_routing_model_shortest_path "Shortest-path" based models: SimGrid calculates shortest
931 paths and manages them. Behaves more or less like most real life
933 2. \ref pf_routing_model_manual "Manually-entered" route models: you have to define all routes
934 manually in the platform description file; this can become
935 tedious very quickly, as it is very verbose.
936 Consistent with some manually managed real life routing.
937 3. \ref pf_routing_model_simple "Simple/fast models": those models offer fast, low memory routing
938 algorithms. You should consider to use this type of model if
939 you can make some assumptions about your AS.
940 Routing in this case is more or less ignored.
942 \subsubsection pf_raf The router affair
944 Using routers becomes mandatory when using shortest-path based
945 models or when using the bindings to the ns-3 packet-level
946 simulator instead of the native analytical network model implemented
949 For graph-based shortest path algorithms, routers are mandatory, because these
950 algorithms require a graph as input and so we need to have source and
951 destination for each edge.
953 Routers are naturally an important concept ns-3 since the
954 way routers run the packet routing algorithms is actually simulated.
955 SimGrid's analytical models however simply aggregate the routing time
956 with the transfer time.
958 So why did we incorporate routers in SimGrid? Rebuilding a graph representation
959 only from the route information turns out to be a very difficult task, because
960 of the missing information about how routes intersect. That is why we
961 introduced routers, which are simply used to express these intersection points.
962 It is important to understand that routers are only used to provide topological
965 To express this topological information, a <b>route</b> has to be
966 defined in order to declare which link is connected to a router.
969 \subsubsection pf_routing_model_shortest_path Shortest-path based models
971 The following table shows all the models that compute routes using
972 shortest-paths algorithms are currently available in SimGrid. More detail on how
973 to choose the best routing model is given in the Section called \"\ref pf_routing_howto_choose_wisely\".
975 | Name | Description |
976 | --------------------------------------------------- | -------------------------------------------------------------------------- |
977 | \ref pf_routing_model_floyd "Floyd" | Floyd routing data. Pre-calculates all routes once |
978 | \ref pf_routing_model_dijkstra "Dijkstra" | Dijkstra routing data. Calculates routes only when needed |
979 | \ref pf_routing_model_dijkstracache "DijkstraCache" | Dijkstra routing data. Handles some cache for already calculated routes. |
981 All those shortest-path models are instanciated in the same way and are
982 completely interchangeable. Here are some examples:
984 \anchor pf_routing_model_floyd
989 <AS id="AS0" routing="Floyd">
991 <cluster id="my_cluster_1" prefix="c-" suffix=""
992 radical="0-1" speed="1000000000" bw="125000000" lat="5E-5"
993 router_id="router1"/>
995 <AS id="AS1" routing="None">
996 <host id="host1" speed="1000000000"/>
999 <link id="link1" bandwidth="100000" latency="0.01"/>
1001 <ASroute src="my_cluster_1" dst="AS1"
1004 <link_ctn id="link1"/>
1010 ASroute given at the end gives a topological information: link1 is
1011 between router1 and host1.
1013 #### Example platform files ####
1015 This is an automatically generated list of example files that use the Floyd
1016 routing model (the path is given relative to SimGrid's source directory)
1018 \verbinclude example_filelist_routing_floyd
1020 \anchor pf_routing_model_dijkstra
1023 #### Example platform files ####
1025 This is an automatically generated list of example files that use the Dijkstra
1026 routing model (the path is given relative to SimGrid's source directory)
1028 \verbinclude example_filelist_routing_dijkstra
1032 <AS id="AS_2" routing="Dijkstra">
1033 <host id="AS_2_host1" speed="1000000000"/>
1034 <host id="AS_2_host2" speed="1000000000"/>
1035 <host id="AS_2_host3" speed="1000000000"/>
1036 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1037 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1038 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1039 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1040 <router id="central_router"/>
1041 <router id="AS_2_gateway"/>
1042 <!-- routes providing topological information -->
1043 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1044 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1045 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1046 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1050 \anchor pf_routing_model_dijkstracache
1051 ### DijkstraCache ###
1053 DijkstraCache example:
1055 <AS id="AS_2" routing="DijkstraCache">
1056 <host id="AS_2_host1" speed="1000000000"/>
1058 (platform unchanged compared to upper example)
1061 #### Example platform files ####
1063 This is an automatically generated list of example files that use the DijkstraCache
1064 routing model (the path is given relative to SimGrid's source directory):
1066 Editor's note: At the time of writing, no platform file used this routing model - so
1067 if there are no example files listed here, this is likely to be correct.
1069 \verbinclude example_filelist_routing_dijkstra_cache
1071 \subsubsection pf_routing_model_manual Manually-entered route models
1073 | Name | Description |
1074 | ---------------------------------- | ------------------------------------------------------------------------------ |
1075 | \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. |
1077 \anchor pf_routing_model_full
1082 <AS id="AS0" routing="Full">
1083 <host id="host1" speed="1000000000"/>
1084 <host id="host2" speed="1000000000"/>
1085 <link id="link1" bandwidth="125000000" latency="0.000100"/>
1086 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
1090 #### Example platform files ####
1092 This is an automatically generated list of example files that use the Full
1093 routing model (the path is given relative to SimGrid's source directory):
1095 \verbinclude example_filelist_routing_full
1097 \subsubsection pf_routing_model_simple Simple/fast models
1099 | Name | Description |
1100 | ---------------------------------------- | ------------------------------------------------------------------------------ |
1101 | \ref pf_routing_model_cluster "Cluster" | This is specific to the \ref pf_tag_cluster "<cluster/>" tag and should not be used by the user, as several assumptions are made. |
1102 | \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. |
1103 | \ref pf_routing_model_vivaldi "Vivaldi" | Perfect when you want to use coordinates. Also see the corresponding \ref pf_P2P_tags "P2P section" below. |
1105 \anchor pf_routing_model_cluster
1109 In this mode, the \ref pf_cabinet "<cabinet/>" tag is available.
1111 #### Example platform files ####
1113 This is an automatically generated list of example files that use the Cluster
1114 routing model (the path is given relative to SimGrid's source directory):
1116 \verbinclude example_filelist_routing_cluster
1118 \anchor pf_routing_model_none
1121 This model does exactly what it's name advertises: Nothing. There is no routing
1122 available within this model and if you try to communicate within the AS that
1123 uses this model, SimGrid will fail unless you have explicitly activated the
1124 \ref options_model_select_network_constant "Constant Network Model" (this model charges
1125 the same for every single communication). It should
1126 be noted, however, that you can still attach an \ref pf_tag_asroute "ASroute",
1127 as is demonstrated in the example below:
1129 \verbinclude platforms/cluster_and_one_host.xml
1131 #### Example platform files ####
1133 This is an automatically generated list of example files that use the None
1134 routing model (the path is given relative to SimGrid's source directory):
1136 \verbinclude example_filelist_routing_none
1139 \anchor pf_routing_model_vivaldi
1142 For more information on how to use the [Vivaldi Coordinates](https://en.wikipedia.org/wiki/Vivaldi_coordinates),
1143 see also Section \ref pf_P2P_tags "P2P tags".
1145 Note that it is possible to combine the Vivaldi routing model with other routing models;
1146 an example can be found in the file \c examples/platforms/cloud.xml. This
1147 examples models a NetZone using Vivaldi that contains other NetZones that use different
1150 #### Example platform files ####
1152 This is an automatically generated list of example files that use the None
1153 routing model (the path is given relative to SimGrid's source directory):
1155 \verbinclude example_filelist_routing_vivaldi
1158 \subsection ps_dec Defining routes
1160 There are currently four different ways to define routes:
1162 | Name | Description |
1163 | ------------------------------------------------- | ----------------------------------------------------------------------------------- |
1164 | \ref pf_tag_route "route" | Used to define route between host/router |
1165 | \ref pf_tag_zoneroute "zoneRoute" | Used to define route between different zones |
1166 | \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. |
1167 | \ref pf_tag_bypassasroute "bypassZoneRoute" | Used in the same way as bypassRoute, but for zones |
1169 Basically all those tags will contain an (ordered) list of references
1170 to link that compose the route you want to define.
1172 Consider the example below:
1175 <route src="Alice" dst="Bob">
1176 <link_ctn id="link1"/>
1177 <link_ctn id="link2"/>
1178 <link_ctn id="link3"/>
1182 The route here from host Alice to Bob will be first link1, then link2,
1183 and finally link3. What about the reverse route? \ref pf_tag_route "Route" and
1184 \ref pf_tag_zoneroute "ASroute" have an optional attribute \c symmetrical, that can
1185 be either \c YES or \c NO. \c YES means that the reverse route is the same
1186 route in the inverse order, and is set to \c YES by default. Note that
1187 this is not the case for bypass*Route, as it is more probable that you
1188 want to bypass only one default route.
1190 For an \ref pf_tag_zoneroute "ASroute", things are just slightly more complicated, as you have
1191 to give the id of the gateway which is inside the AS you want to access ...
1192 So it looks like this:
1195 <ASroute src="AS1" dst="AS2"
1196 gw_src="router1" gw_dst="router2">
1197 <link_ctn id="link1"/>
1201 gw == gateway, so when any message are trying to go from AS1 to AS2,
1202 it means that it must pass through router1 to get out of the AS, then
1203 pass through link1, and get into AS2 by being received by router2.
1204 router1 must belong to AS1 and router2 must belong to AS2.
1206 \subsubsection pf_tag_linkctn <link_ctn>
1208 This entity has only one purpose: Refer to an already existing
1209 \ref pf_tag_link "<link/>" when defining a route, i.e., it
1210 can only occur as a child of \ref pf_tag_route "<route/>"
1212 | Attribute name | Mandatory | Values | Description |
1213 | --------------- | --------- | ------ | ----------- |
1214 | id | yes | String | The identifier of the link that should be added to the route. |
1215 | 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.
1217 #### Example Files ####
1219 This is an automatically generated list of example files that use the \c <link_ctn/>
1220 entity (the path is given relative to SimGrid's source directory):
1222 \verbinclude example_filelist_xmltag_linkctn
1224 \subsubsection pf_tag_zoneroute <zoneRoute>
1226 The purpose of this entity is to define a route between two ASes.
1227 This is mainly useful when you're in the \ref pf_routing_model_full "Full routing model".
1229 #### Attributes ####
1231 | Attribute name | Mandatory | Values | Description |
1232 | --------------- | --------- | ------ | ----------- |
1233 | src | yes | String | The identifier of the source AS |
1234 | dst | yes | String | See the \c src attribute |
1235 | 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. |
1236 | gw_dst | yes | String | Same as \c gw_src, but with the dst AS instead. |
1237 | symmetrical | no | YES\|NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1242 <AS id="AS0" routing="Full">
1243 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
1244 radical="0-149" speed="1000000000" bw="125000000" lat="5E-5"
1245 bb_bw="2250000000" bb_lat="5E-4"/>
1247 <cluster id="my_cluster_2" prefix="c-" suffix=".me"
1248 radical="150-299" speed="1000000000" bw="125000000" lat="5E-5"
1249 bb_bw="2250000000" bb_lat="5E-4"/>
1251 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1253 <ASroute src="my_cluster_1" dst="my_cluster_2"
1254 gw_src="c-my_cluster_1_router.me"
1255 gw_dst="c-my_cluster_2_router.me">
1256 <link_ctn id="backbone"/>
1258 <ASroute src="my_cluster_2" dst="my_cluster_1"
1259 gw_src="c-my_cluster_2_router.me"
1260 gw_dst="c-my_cluster_1_router.me">
1261 <link_ctn id="backbone"/>
1266 \subsubsection pf_tag_route <route>
1268 The principle is the same as for
1269 \ref pf_tag_zoneroute "ASroute": The route contains a list of links that
1270 provide a path from \c src to \c dst. Here, \c src and \c dst can both be either a
1271 \ref pf_tag_host "host" or \ref pf_router "router". This is mostly useful for the
1272 \ref pf_routing_model_full "Full routing model" as well as for the
1273 \ref pf_routing_model_shortest_path "shortest-paths" based models (as they require
1274 topological information).
1277 | Attribute name | Mandatory | Values | Description |
1278 | --------------- | --------- | ---------------------- | ----------- |
1279 | src | yes | String | The value given to the source's "id" attribute |
1280 | dst | yes | String | The value given to the destination's "id" attribute. |
1281 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1286 A route in the \ref pf_routing_model_full "Full routing model" could look like this:
1288 <route src="Tremblay" dst="Bourassa">
1289 <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"/>
1293 A route in the \ref pf_routing_model_shortest_path "Shortest-Path routing model" could look like this:
1295 <route src="Tremblay" dst="Bourassa">
1300 You must only have one link in your routes when you're using them to provide
1301 topological information, as the routes here are simply the edges of the
1302 (network-)graph and the employed algorithms need to know which edge connects
1303 which pair of entities.
1305 \subsubsection pf_tag_bypassasroute bypassASroute
1307 As said before, once you choose
1308 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1309 define some of your routes, which will be specific. You may also want
1310 to bypass some routes defined in lower level AS at an upper stage:
1311 <b>bypassASroute</b> is the tag you're looking for. It allows to
1312 bypass routes defined between already defined between AS (if you want
1313 to bypass route for a specific host, you should just use byPassRoute).
1314 The principle is the same as ASroute : <b>bypassASroute</b> contains
1315 list of links that are in the path between src and dst.
1317 #### Attributes ####
1319 | Attribute name | Mandatory | Values | Description |
1320 | --------------- | --------- | ---------------------- | ----------- |
1321 | src | yes | String | The value given to the source AS's "id" attribute |
1322 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1323 | 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 |
1324 | 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|
1325 | symmetrical | no | YES\| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1330 <bypassASRoute src="my_cluster_1" dst="my_cluster_2"
1331 gw_src="my_cluster_1_router"
1332 gw_dst="my_cluster_2_router">
1333 <link_ctn id="link_tmp"/>
1337 This example shows that link \c link_tmp (definition not displayed here) directly
1338 connects the router \c my_cluster_1_router in the source cluster to the router
1339 \c my_cluster_2_router in the destination router. Additionally, as the \c symmetrical
1340 attribute was not given, this route is presumed to be symmetrical.
1342 \subsubsection pf_tag_bypassroute bypassRoute
1344 As said before, once you choose
1345 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1346 define some of your routes, which will be specific. You may also want
1347 to bypass some routes defined in lower level AS at an upper stage :
1348 <b>bypassRoute</b> is the tag you're looking for. It allows to bypass
1349 routes defined between <b>host/router</b>. The principle is the same
1350 as route : <b>bypassRoute</b> contains list of links references of
1351 links that are in the path between src and dst.
1353 #### Attributes ####
1355 | Attribute name | Mandatory | Values | Description |
1356 | --------------- | --------- | ---------------------- | ----------- |
1357 | src | yes | String | The value given to the source AS's "id" attribute |
1358 | dst | yes | String | The value given to the destination AS's "id" attribute. |
1359 | symmetrical | no | YES \| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1364 <bypassRoute src="host_1" dst="host_2">
1365 <link_ctn id="link_tmp"/>
1369 This example shows that link \c link_tmp (definition not displayed here) directly
1370 connects host \c host_1 to host \c host_2. Additionally, as the \c symmetrical
1371 attribute was not given, this route is presumed to be symmetrical.
1373 \subsection pb_baroex Basic Routing Example
1375 Let's say you have an AS named AS_Big that contains two other AS, AS_1
1376 and AS_2. If you want to make a host (h1) from AS_1 with another one
1377 (h2) from AS_2 then you'll have to proceed as follows:
1378 \li First, you have to ensure that a route is defined from h1 to the
1379 AS_1's exit gateway and from h2 to AS_2's exit gateway.
1380 \li Then, you'll have to define a route between AS_1 to AS_2. As those
1381 AS are both resources belonging to AS_Big, then it has to be done
1382 at AS_big level. To define such a route, you have to give the
1383 source AS (AS_1), the destination AS (AS_2), and their respective
1384 gateway (as the route is effectively defined between those two
1385 entry/exit points). Elements of this route can only be elements
1386 belonging to AS_Big, so links and routers in this route should be
1387 defined inside AS_Big. If you choose some shortest-path model,
1388 this route will be computed automatically.
1390 As said before, there are mainly 2 tags for routing :
1391 \li <b>ASroute</b>: to define routes between two <b>AS</b>
1392 \li <b>route</b>: to define routes between two <b>host/router</b>
1394 As we are dealing with routes between AS, it means that those we'll
1395 have some definition at AS_Big level. Let consider AS_1 contains 1
1396 host, 1 link and one router and AS_2 3 hosts, 4 links and one router.
1397 There will be a central router, and a cross-like topology. At the end
1398 of the crosses arms, you'll find the 3 hosts and the router that will
1399 act as a gateway. We have to define routes inside those two AS. Let
1400 say that AS_1 contains full routes, and AS_2 contains some Floyd
1401 routing (as we don't want to bother with defining all routes). As
1402 we're using some shortest path algorithms to route into AS_2, we'll
1403 then have to define some <b>route</b> to gives some topological
1404 information to SimGrid. Here is a file doing it all :
1407 <AS id="AS_Big" routing="Dijkstra">
1408 <AS id="AS_1" routing="Full">
1409 <host id="AS_1_host1" speed="1000000000"/>
1410 <link id="AS_1_link" bandwidth="1250000000" latency="5E-4"/>
1411 <router id="AS_1_gateway"/>
1412 <route src="AS_1_host1" dst="AS_1_gateway">
1413 <link_ctn id="AS_1_link"/>
1416 <AS id="AS_2" routing="Floyd">
1417 <host id="AS_2_host1" speed="1000000000"/>
1418 <host id="AS_2_host2" speed="1000000000"/>
1419 <host id="AS_2_host3" speed="1000000000"/>
1420 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1421 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1422 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1423 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1424 <router id="central_router"/>
1425 <router id="AS_2_gateway"/>
1426 <!-- routes providing topological information -->
1427 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1428 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1429 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1430 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1432 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1434 <ASroute src="AS_1" dst="AS_2"
1435 gw_src="AS_1_gateway"
1436 gw_dst="AS_2_gateway">
1437 <link_ctn id="backbone"/>
1442 \section pf_other Other tags
1444 The following tags can be used inside a \<platform\> tag even if they are not
1445 directly describing the platform:
1447 - @ref pf_tag_config passes configuration options, e.g. to change the network model;
1448 - @ref pf_tag_prop gives user-defined properties to various elements
1450 \subsection pf_tag_config <config>
1452 Adding configuration flags into the platform file is particularly
1453 useful when the described platform is best used with specific
1454 flags. For example, you could finely tune SMPI in your platform file directly.
1456 | Attribute | Values | Description |
1457 | ---------- | ------------------- | ---------------------------------------------- |
1458 | id | String (optional) | This optional identifier is ignored by SimGrid |
1460 * **Included tags:** @ref pf_tag_prop to specify a given configuration item (see @ref options).
1462 Any such configuration must be given at the very top of the platform file.
1467 <?xml version='1.0'?>
1468 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1469 <platform version="4">
1471 <prop id="maxmin/precision" value="0.000010" />
1472 <prop id="cpu/optim" value="TI" />
1473 <prop id="network/model" value="SMPI" />
1474 <prop id="smpi/bw-factor" value="65472:0.940694;15424:0.697866;9376:0.58729" />
1477 <AS id="AS0" routing="Full">
1481 \subsection pf_tag_prop <prop>
1483 Defines a user-defined property, identified with a name and having a
1484 value. You can specify such properties to most kind of resources:
1485 @ref pf_tag_zone, @ref pf_tag_host, @ref pf_tag_storage,
1486 @ref pf_tag_cluster and @ref pf_tag_link. These values can be retrieved
1487 at runtime with MSG_zone_property() or simgrid::s4u::NetZone::property(),
1488 or similar functions.
1490 | Attribute | Values | Description |
1491 | --------- | ---------------------- | ----------------------------------------------------------------------------------------- |
1492 | id | String (mandatory) | Identifier of this property. Must be unique for a given property holder, eg host or link. |
1493 | value | String (mandatory) | Value of this property; The semantic is completely up to you. |
1495 * **Included tags:** none.
1500 <prop id="Operating System" value="Linux" />
1504 \subsection pf_trace trace and trace_connect
1506 Both tags are an alternate way to pass files containing information on
1507 availability, state etc. to an entity. (See also, for instance, Section \ref
1508 pf_host_churn "Churn", as described for the host entity.) Instead of referring
1509 to the file directly in the host, link, or cluster tag, you proceed by defining
1510 a trace with an id corresponding to a file, later a host/link/cluster, and
1511 finally using trace_connect you say that the file trace must be used by the
1518 <AS id="AS0" routing="Full">
1519 <host id="bob" speed="1000000000"/>
1521 <trace id="myTrace" file="bob.trace" periodicity="1.0"/>
1522 <trace_connect trace="myTrace" element="bob" kind="POWER"/>
1526 The order here is important. \c trace_connect must come
1527 after the elements \c trace and \c host, as both the host
1528 and the trace definition must be known when \c trace_connect
1529 is parsed; the order of \c trace and \c host is arbitrary.
1532 #### \c trace attributes ####
1535 | Attribute name | Mandatory | Values | Description |
1536 | --------------- | --------- | ---------------------- | ----------- |
1537 | id | yes | String | Identifier of this trace; this is the name you pass on to \c trace_connect. |
1538 | 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. |
1539 | trace_periodicity | yes | String | This is the same as for \ref pf_tag_host "hosts" (see there for details) |
1541 Here is an example of trace when no file name is provided:
1544 <trace id="myTrace" periodicity="1.0">
1551 #### \c trace_connect attributes ####
1553 | Attribute name | Mandatory | Values | Description |
1554 | --------------- | --------- | ---------------------- | ----------- |
1555 | kind | no | HOST_AVAIL\|POWER\|<br/>LINK_AVAIL\|BANDWIDTH\|LATENCY (Default: HOST_AVAIL) | Describes the kind of trace. |
1556 | trace | yes | String | Identifier of the referenced trace (specified of the trace's \c id attribute) |
1557 | element | yes | String | The identifier of the referenced entity as given by its \c id attribute |
1559 \section pf_hints Hints, tips and frequently requested features
1561 Now you should know at least the syntax and be able to create a
1562 platform by your own. However, after having ourselves wrote some platforms, there
1563 are some best practices you should pay attention to in order to
1564 produce good platform and some choices you can make in order to have
1565 faster simulations. Here's some hints and tips, then.
1567 @subsection pf_hints_search Finding the platform example that you need
1569 Most platform files that we ship are in the @c examples/platforms
1570 folder. The good old @c grep tool can find the examples you need when
1571 wondering on a specific XML tag. Here is an example session searching
1572 for @ref pf_trace "trace_connect":
1575 % cd examples/platforms
1576 % grep -R -i -n --include="*.xml" "trace_connect" .
1577 ./two_hosts_platform_with_availability_included.xml:26:<trace_connect kind="SPEED" trace="A" element="Cpu A"/>
1578 ./two_hosts_platform_with_availability_included.xml:27:<trace_connect kind="HOST_AVAIL" trace="A_failure" element="Cpu A"/>
1579 ./two_hosts_platform_with_availability_included.xml:28:<trace_connect kind="SPEED" trace="B" element="Cpu B"/>
1580 ./two_hosts.xml:17: <trace_connect trace="Tremblay_power" element="Tremblay" kind="SPEED"/>
1583 \subsection pf_as_h AS Hierarchy
1584 The AS design allows SimGrid to go fast, because computing route is
1585 done only for the set of resources defined in this AS. If you're using
1586 only a big AS containing all resource with no AS into it and you're
1587 using Full model, then ... you'll loose all interest into it. On the
1588 other hand, designing a binary tree of AS with, at the lower level,
1589 only one host, then you'll also loose all the good AS hierarchy can
1590 give you. Remind you should always be "reasonable" in your platform
1591 definition when choosing the hierarchy. A good choice if you try to
1592 describe a real life platform is to follow the AS described in
1593 reality, since this kind of trade-off works well for real life
1596 \subsection pf_exit_as Exit AS: why and how
1597 Users that have looked at some of our platforms may have notice a
1598 non-intuitive schema ... Something like that :
1602 <AS id="AS_4" routing="Full">
1603 <AS id="exitAS_4" routing="Full">
1604 <router id="router_4"/>
1606 <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"/>
1607 <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"/>
1608 <link id="4_1" bandwidth="2250000000" latency="5E-5"/>
1609 <link id="4_2" bandwidth="2250000000" latency="5E-5"/>
1610 <link id="bb_4" bandwidth="2250000000" latency="5E-4"/>
1611 <ASroute src="cl_4_1"
1613 gw_src="c_4_1-cl_4_1_router"
1614 gw_dst="c_4_2-cl_4_2_router">
1615 <link_ctn id="4_1"/>
1616 <link_ctn id="bb_4"/>
1617 <link_ctn id="4_2"/>
1619 <ASroute src="cl_4_1"
1621 gw_src="c_4_1-cl_4_1_router"
1623 <link_ctn id="4_1"/>
1624 <link_ctn id="bb_4"/>
1626 <ASroute src="cl_4_2"
1628 gw_src="c_4_2-cl_4_2_router"
1630 <link_ctn id="4_2"/>
1631 <link_ctn id="bb_4"/>
1636 In the AS_4, you have an exitAS_4 defined, containing only one router,
1637 and routes defined to that AS from all other AS (as cluster is only a
1638 shortcut for an AS, see cluster description for details). If there was
1639 an upper AS, it would define routes to and from AS_4 with the gateway
1640 router_4. It's just because, as we did not allowed (for performances
1641 issues) to have routes from an AS to a single host/router, you have to
1642 enclose your gateway, when you have AS included in your AS, within an
1643 AS to define routes to it.
1645 \subsection pf_P2P_tags P2P or how to use coordinates
1646 SimGrid allows you to use some coordinated-based system, like vivaldi,
1647 to describe a platform. The main concept is that you have some peers
1648 that are located somewhere: this is the function of the
1649 <b>coordinates</b> of the \<peer\> or \<host\> tag. There's nothing
1650 complicated in using it, here is an example:
1653 <?xml version='1.0'?>
1654 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1655 <platform version="4">
1657 <AS id="AS0" routing="Vivaldi">
1658 <host id="100030591" coordinates="25.5 9.4 1.4" speed="1.5Gf" />
1659 <host id="100036570" coordinates="-12.7 -9.9 2.1" speed="7.3Gf" />
1661 <host id="100429957" coordinates="17.5 6.7 18.8" speed="8.3Gf" />
1666 Coordinates are then used to calculate latency (in microseconds)
1667 between two hosts by calculating the distance between the two hosts
1668 coordinates with the following formula: distance( (x1, y1, z1), (x2,
1669 y2, z2) ) = euclidian( (x1,y1), (x2,y2) ) + abs(z1) + abs(z2)
1671 In other words, we take the euclidian distance on the two first
1672 dimensions, and then add the absolute values found on the third
1673 dimension. This may seem strange, but it was found to allow better
1674 approximations of the latency matrices (see the paper describing
1677 Note that the previous example defines a routing directly between hosts but it could be also used to define a routing between AS.
1678 That is for example what is commonly done when using peers (see Section \ref pf_peer).
1680 <?xml version='1.0'?>
1681 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1682 <platform version="4">
1684 <AS id="AS0" routing="Vivaldi">
1685 <peer id="peer-0" coordinates="173.0 96.8 0.1" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us"/>
1686 <peer id="peer-1" coordinates="247.0 57.3 0.6" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1687 <peer id="peer-2" coordinates="243.4 58.8 1.4" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1691 In such a case though, we connect the AS created by the <b>peer</b> tag with the Vivaldi routing mechanism.
1692 This means that to route between AS1 and AS2, it will use the coordinates of router_AS1 and router_AS2.
1693 This is currently a convention and we may offer to change this convention in the DTD later if needed.
1694 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.
1697 \subsection pf_routing_howto_choose_wisely Choosing wisely the routing model to use
1700 Choosing wisely the routing model to use can significantly fasten your
1701 simulation/save your time when writing the platform/save tremendous
1702 disk space. Here is the list of available model and their
1703 characteristics (lookup : time to resolve a route):
1705 \li <b>Full</b>: Full routing data (fast, large memory requirements,
1707 \li <b>Floyd</b>: Floyd routing data (slow initialization, fast
1708 lookup, lesser memory requirements, shortest path routing only).
1709 Calculates all routes at once at the beginning.
1710 \li <b>Dijkstra</b>: Dijkstra routing data (fast initialization, slow
1711 lookup, small memory requirements, shortest path routing only).
1712 Calculates a route when necessary.
1713 \li <b>DijkstraCache</b>: Dijkstra routing data (fast initialization,
1714 fast lookup, small memory requirements, shortest path routing
1715 only). Same as Dijkstra, except it handles a cache for latest used
1717 \li <b>None</b>: No routing (usable with Constant network only).
1718 Defines that there is no routes, so if you try to determine a
1719 route without constant network within this AS, SimGrid will raise
1721 \li <b>Vivaldi</b>: Vivaldi routing, so when you want to use coordinates
1722 \li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
1725 \subsection pf_switch I want to describe a switch but there is no switch tag!
1727 Actually we did not include switch tag. But when you're trying to
1728 simulate a switch, assuming
1729 fluid bandwidth models are used (which SimGrid uses by default unless
1730 ns-3 or constant network models are activated), the limiting factor is
1731 switch backplane bandwidth. So, essentially, at least from
1732 the simulation perspective, a switch is similar to a
1733 link: some device that is traversed by flows and with some latency and
1734 so,e maximum bandwidth. Thus, you can simply simulate a switch as a
1736 can be connected to this "switch", which is then included in routes just
1740 \subsection pf_multicabinets I want to describe multi-cabinets clusters!
1742 You have several possibilities, as usual when modeling things. If your
1743 cabinets are homogeneous and the intercabinet network negligible for
1744 your study, you should just create a larger cluster with all hosts at
1747 In the rare case where your hosts are not homogeneous between the
1748 cabinets, you can create your cluster completely manually. For that,
1749 create an As using the Cluster routing, and then use one
1750 <cabinet> for each cabinet. This cabinet tag can only be used an
1751 As using the Cluster routing schema, and creating
1753 Be warned that creating a cluster manually from the XML with
1754 <cabinet>, <backbone> and friends is rather tedious. The
1755 easiest way to retrieve some control of your model without diving into
1756 the <cluster> internals is certainly to create one separate
1757 <cluster> per cabinet and interconnect them together. This is
1758 what we did in the G5K example platform for the Graphen cluster.
1760 \subsection pf_platform_multipath I want to express multipath routing in platform files!
1762 It is unfortunately impossible to express the fact that there is more
1763 than one routing path between two given hosts. Let's consider the
1764 following platform file:
1767 <route src="A" dst="B">
1770 <route src="B" dst="C">
1773 <route src="A" dst="C">
1778 Although it is perfectly valid, it does not mean that data traveling
1779 from A to C can either go directly (using link 3) or through B (using
1780 links 1 and 2). It simply means that the routing on the graph is not
1781 trivial, and that data do not following the shortest path in number of
1782 hops on this graph. Another way to say it is that there is no implicit
1783 in these routing descriptions. The system will only use the routes you
1784 declare (such as <route src="A" dst="C"><link_ctn
1785 id="3"/></route>), without trying to build new routes by aggregating
1788 You are also free to declare platform where the routing is not
1789 symmetrical. For example, add the following to the previous file:
1792 <route src="C" dst="A">
1798 This makes sure that data from C to A go through B where data from A
1799 to C go directly. Don't worry about realism of such settings since
1800 we've seen ways more weird situation in real settings (in fact, that's
1801 the realism of very regular platforms which is questionable, but
1802 that's another story).
