1 /*! \page platform Platform Description
5 In order to run any simulation, SimGrid needs 3 things: something to run
6 (so, your code), a description of the platform on which you want to run your
7 application, and finally it needs something to know where to deploy what.
9 For the latest 2 entries, you have basically 2 ways to give it as an input :
10 \li You can program it, either using the Lua console (\ref
11 MSG_Lua_funct) or if you're using MSG some of its platform and
12 deployments functions(\ref msg_simulation). If you want to use it,
13 please refer to its doc. (you can also check the section \ref
14 pf_flexml_bypassing but this is strongly deprecated, as there is a
15 new way to do it properly, but not yet documented).
16 \li You can use two XML files: a platform description file and a
17 deployment description one.
19 For the deployment stuff, please takea look at \ref deployment
21 The platform description may be complicated. This documentation is all
22 about how to write this file: what are the basic concept it relies on,
23 what possibilities are offered, and some hints and tips on how to
24 write a good platform description.
26 \section pf_overview Some words about XML and DTD
28 We choose to use XML because of some of its possibilities: if you're
29 using an accurate XML editor, or simply using any XML plug-in for
30 eclipse, it will allow you to have cool stuff like auto-completion,
31 validation and checking, so all syntaxic errors may be avoided this
34 the XML checking is done based on the dtd which is nowaday online at
35 <a href="http://simgrid.gforge.inria.fr/simgrid.dtd">http://simgrid.gforge.inria.fr/simgrid.dtd</a>
36 while you might be tempted to read it, it will not help you that much.
38 If you read it, you should notice two or three important things :
39 \li The platform tags contains a version attributes. At the time of
40 writing this doc the current version is 3.
41 \li The DTD contains definitions for the 2 files used by SimGrid (platform
42 description and deployment).
43 \li There is a bunch of possibilities ! Let's see what's in it
46 \section pf_basics Basic concepts
48 Nowadays, the Internet is composed of a bunch of independently managed
49 networks. Within each of those networks, there are entry and exit
50 points (most of the time, you can both enter and exit through the same
51 point) that allows to go out of the current network and reach other
52 networks. At the upper level, these networks are known as
53 <b>Autonomous System (AS)</b>, while at the lower level they are named
54 sub-networks, or LAN. Indeed they are autonomous: routing is defined
55 within the limits of his network by the administrator, and so, those
56 networks can continue to operate without the existence of other
57 networks. There are some rules to get out of networks by the entry
58 points (or gateways). Those gateways allow you to go from a network to
59 another one. Inside of each autonomous system, there is a bunch of
60 equipments (cables, routers, switches, computers) that belong to the
61 autonomous system owner.
63 SimGrid platform description file relies exactly on the same concepts
64 as real life platform. Every resource (computers, network equipments,
65 and so on) belongs to an AS. Within this AS, you can define the
66 routing you want between its elements (that's done with the routing
67 model attribute and eventually with some \<route\> tag). You define AS
68 by using ... well ... the \<AS\> tag. An AS can also contain some AS :
69 AS allows you to define the hierarchy of your platform.
71 Within each AS, you basically have the following type of resources:
72 \li <b>host</b>: an host, with cores in it, and so on
73 \li <b>router</b>: a router or a gateway.
74 \li <b>link</b>: a link, that defines a connection between two (or
75 more) resources (and have a bandwidth and a latency)
76 \li <b>cluster</b>: like a real cluster, contains many hosts
77 interconnected by some dedicated network.
79 Between those elements, a routing has to be defined. As the AS is
80 supposed to be Autonomous, this has to be done at the AS level. As AS
81 handles two different types of entities (<b>host/router</b> and
82 <b>AS</b>) you will have to define routes between those elements. A
83 network model have to be provided for AS, but you may/will need,
84 depending of the network model, or because you want to bypass the
85 default beahviour to defines routes manually. There are 3 tags to use:
86 \li <b>ASroute</b>: to define routes between two <b>AS</b>
87 \li <b>route</b>: to define routes between two <b>host/router</b>
88 \li <b>bypassRoute</b>: to define routes between two <b>AS</b> that
89 will bypass default routing.
91 Here is an illustration of the overall concepts:
94 <a href="AS_hierarchy.png" border=0><img src="AS_hierarchy.png" width="30%" border=0 align="center"></a>
97 Circles represent processing units and squares represent network routers. Bold
98 lines represent communication links. AS2 models the core of a national
99 network interconnecting a small flat cluster (AS4) and a larger
100 hierarchical cluster (AS5), a subset of a LAN (AS6), and a set of peers
101 scattered around the world (AS7).
104 This is all for the concepts ! To make a long story short, a SimGrid
105 platform is made of a hierarchy of AS, each of them containing
106 resources, and routing is defined at AS level. Let's have a deeper
111 \section pf_pftags Describing resources and their organization
113 \subsection pf_As Platform organization tag : AS
115 AS (or Autonomous System) is an organizational unit that contains
116 resources and defines routing between them, and eventually some other
117 AS. So it allows you to define a hierarchy into your platform.
118 <b>*ANY*</b> resource <b>*MUST*</b> belong to an AS. There are a few
121 <b>AS</b> attributes :
122 \li <b>id (mandatory)</b>: the identifier of AS to be used when
124 \li <b>routing (mandatory)</b>: the routing model used into it. By
125 model we mean the internal way the simulator will manage routing.
126 That also have a big impact on how many information you'll have to
127 provide to help the simulator to route between the AS elements.
128 <b>routing</b> possible values are <b>Full, Floyd, Dijkstra,
129 DijkstraCache, none, RuleBased, Vivaldi, Cluster</b>. For more
130 explanation about what to choose, take a look at the section
133 Elements into an AS are basically resources (computers, network
134 equipments) and some routing informations if necessary (see below for
139 <AS id="AS0" routing="Full">
140 <host id="host1" power="1000000000"/>
141 <host id="host2" power="1000000000"/>
142 <link id="link1" bandwidth="125000000" latency="0.000100"/>
143 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
147 In this example, AS0 contains two hosts (host1 and host2). The route
148 between the hosts goes through link1.
151 \subsection pf_Cr Computing resources: hosts, clusters and peers.
153 \subsubsection pf_host host
155 A <b>host</b> represents a computer, where you will be able to execute
156 code and from which you can send and receive information. A host can
157 contain more than 1 core. Here are the attributes of a host :
160 <b>host</b> attributes :
161 \li <b>id (mandatory)</b>: the identifier of the host to be used when
163 \li <b>power (mandatory)</b>:the peak number FLOPS the CPU can manage.
165 \li <b>core</b>: The number of core of this host. If setted, the power
166 gives the power of one core. The specified computing power will be
167 available to up to 6 sequential tasks without sharing. If more
168 tasks are placed on this host, the resource will be shared
169 accordingly. For example, if you schedule 12 tasks on the host,
170 each will get half of the computing power. Please note that
171 although sound, this model were never scientifically assessed.
172 Please keep this fact in mind when using it.
174 \li <b>availability</b>: specify if the percentage of power available.
175 \li <b>availability_file</b>: Allow you to use a file as input. This
176 file will contain availability traces for this computer. The
177 syntax of this file is defined below. Possible values : absolute
178 or relative path, syntax similar to the one in use on your system.
179 \li <b>state</b>: the computer state, as in : is that computer ON or
180 OFF. Possible values : "ON" or "OFF".
181 \li <b>state_file</b>: Same mechanism as availability_file, similar
183 \li <b>coordinates</b>: you'll have to give it if you choose the
184 vivaldi, coordinate-based routing model for the AS the host
185 belongs to. More details about it in the P2P coordinate based
188 An host can contain some <b>mount</b> that defines mounting points
189 between some storage resource and the <b>host</b>. Please refer to the
190 storage doc for more information.
192 An host can also contain the <b>prop</b> tag. the prop tag allows you
193 to define additional informations on this host following the
194 attribute/value schema. You may want to use it to give information to
195 the tool you use for rendering your simulation, for example.
199 <host id="host1" power="1000000000"/>
200 <host id="host2" power="1000000000">
201 <prop id="color" value="blue"/>
202 <prop id="rendershape" value="square"/>
207 <b>Expressing dynamicity.</b>
208 It is also possible to seamlessly declare a host whose
209 availability changes over time using the availability_file
210 attribute and a separate text file whose syntax is exemplified below.
212 <b>Adding a trace file</b>
214 <platform version="1">
215 <host id="bob" power="500000000"
216 availability_file="bob.trace" />
219 <b>Example of "bob.trace" file</b>
227 At time 0, our host will deliver 500~Mflop/s. At time 11.0, it will
228 deliver half, that is 250~Mflop/s until time 20.0 where it will
229 will start delivering 80\% of its power, that is 400~Mflop/s. Last, at
230 time 21.0 (20.0 plus the periodicity 1.0), we loop back to the
231 beginning and the host will deliver again 500~Mflop/s.
233 <b>Changing initial state</b>
235 It is also possible to specify whether the host
236 is up or down by setting the <b>state</b> attribute to either <b>ON</b>
237 (default value) or <b>OFF</b>.
239 <b>Expliciting the default value "ON"</b>
241 <platform version="1">
247 <b>Host switched off</b>
249 <platform version="1">
255 <b>Expressing churn</b>
256 To express the fact that a host can change state over time (as in P2P
257 systems, for instance), it is possible to use a file describing the time
258 at which the host is turned on or off. An example of the content
259 of such a file is presented below.
260 <b>Adding a state file</b>
262 <platform version="1">
263 <host id="bob" power="500000000"
264 state_file="bob.fail" />
267 <b>Example of "bob.fail" file</b>
274 A negative value means <b>down</b> while a positive one means <b>up and
275 running</b>. From time 0.0 to time 1.0, the host is on. At time 1.0, it is
276 turned off and at time 2.0, it is turned on again until time 12 (2.0 plus the
277 periodicity 10.0). It will be turned on again at time 13.0 until time 23.0, and
282 \subsubsection pf_cluster cluster
284 A <b>cluster</b> represents a cluster. It is most of the time used
285 when you want to have a bunch of machine defined quickly. It must be
286 noted that cluster is meta-tag : <b>from the inner SimGrid point of
287 view, a cluster is an AS where some optimized routing is defined</b>.
288 The default inner organisation of the cluster is as follow:
294 ____________|__________|_____________ backbone
296 l0| l1| l2| l97| l96 | | l99
302 You have a set of <b>host</b> defined. Each of them has a <b>link</b>
303 to a central backbone (backbone is a <b>link</b> itsef, as a link can
304 be used to represent a switch, see the switch or <b>link</b> section
305 below for more details about it). A <b>router</b> gives a way to the
306 <b>cluster</b> to be connected to the outside world. Internally,
307 cluster is then an AS containing all hosts : the router is the default
308 gateway for the cluster.
310 There is an alternative organization, which is as follow :
324 The principle is the same, except we don't have the backbone. The way
325 to obtain it is simple : you just have to let bb_* attributes
330 <b>cluster</b> attributes :
331 \li <b>id (mandatory)</b>: the identifier of the cluster to be used
332 when referring to it.
333 \li <b>prefix (mandatory)</b>: each node of the cluster has to have a
334 name. This is its prefix.
335 \li <b>suffix (mandatory)</b>: node suffix name.
336 \li <b>radical (mandatory)</b>: regexp used to generate cluster nodes
337 name. Syntax is quite common, "10-20" will give you 11 machines
338 numbered from 10 to 20, "10-20;2" will give you 12 machines, one
339 with the number 2, others numbered as before. The produced number
340 is concatenated between prefix and suffix to form machine names.
341 \li <b>power (mandatory)</b>: same as <b>host</b> power.
342 \li <b>core</b>: same as <b>host</b> core.
343 \li <b>bw (mandatory)</b>: bandwidth for the links between nodes and
344 backbone (if any). See <b>link</b> section for syntax/details.
345 \li <b>lat (mandatory)</b>: latency for the links between nodes and
346 backbone (if any). See <b>link</b> section for syntax/details.
347 \li <b>sharing_policy</b>: sharing policy for the links between nodes
348 and backbone (if any). See <b>link</b> section for syntax/details.
349 \li <b>bb_bw </b>: bandwidth for backbone (if any). See <b>link</b>
350 section for syntax/details. If both bb_* attributes are ommited,
351 no backbone is created (alternative cluster architecture described
353 \li <b>bb_lat </b>: latency for backbone (if any). See <b>link</b>
354 section for syntax/details. If both bb_* attributes are ommited,
355 no backbone is created (alternative cluster architecture described
357 \li <b>bb_sharing_policy</b>: sharing policy for the backbone (if
358 any). See <b>link</b> section for syntax/details.
359 \li <b>availability_file</b>: Allow you to use a file as input for
360 availability. Similar to <b>hosts</b> attribute.
361 \li <b>state_file</b>: Allow you to use a file as input for states.
362 Similar to <b>hosts</b> attribute.
364 the router name is defined as the resulting String in the following
368 router_name = prefix + clusterId + router_ + suffix;
372 <b>cluster example</b>
374 <cluster id="my_cluster_1" prefix="" suffix=""
375 radical="0-262144" power="1000000000" bw="125000000" lat="5E-5"/>
376 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
377 radical="0-99" power="1000000000" bw="125000000" lat="5E-5"
378 bb_bw="2250000000" bb_lat="5E-4"/>
380 The second examples creates 100 machines, which names are the following:
389 \subsubsection pf_peer peer
390 A <b>peer</b> represents a peer, as in Peer-to-Peer (P2P). Basically,
391 as cluster, <b>A PEER IS INTERNALLY INTERPRETED AS AN \<AS\></b>. It's
392 just a kind of shortcut that does the following :
394 \li It creates an host
395 \li Two links : one for download and one for upload. This is
396 convenient to use and simulate stuff under the last mile model (as
398 \li It creates a gateway that serve as entry point for this peer zone.
399 This router has coordinates.
401 <b>peer</b> attributes :
402 \li <b>id (mandatory)</b>: the identifier of the peer to be used when
404 \li <b>power CDATA (mandatory)</b>: as in host
405 \li <b>bw_in CDATA (mandatory)</b>: bandwidth in.
406 \li <b>bw_out CDATA (mandatory)</b>:bandwidth out.
407 \li <b>lat CDATA (mandatory)</b>: Latency for in and out links.
408 \li <b>coordinates</b>: coordinates of the gateway for this peer.
409 \li <b>sharing_policy</b>: sharing policy for links. Can be SHARED or
410 FULLDUPLEX, FULLDUPLEX is the default. See <b>link</b> description
412 \li <b>availability_file</b>: availability file for the peer. Same as
413 host availability file. See <b>host</b> description for details.
414 \li <b>state_file </b>: state file for the peer. Same as host state
415 file. See <b>host</b> description for details.
417 \subsection pf_ne Network equipments: links and routers
419 You have basically two entities available to represent network entities:
420 \li <b>link</b>: represents something that has a limited bandwidth, a
421 latency, and that can be shared according to TCP way to share this
422 bandwidth. <b>LINKS ARE NOT EDGES BUT HYPEREDGES</b>: it means
423 that you can have more than 2 equipments connected to it.
424 \li <b>router</b>: represents something that one message can be routed
425 to, but does not accept any code, nor have any influence on the
426 performances (no bandwidth, no latency, not anything).<b>ROUTERS
427 ARE ENTITIES (ALMOST) IGNORED BY THE SIMULATOR WHEN THE SIMULATION
428 HAS BEGUN</b>. If you want to represent something like a switch,
429 you must use <b>link</b> (see section below). Routers are used in
430 order to run some routing algorithm and determine routes (see
431 routing section for details).
433 let's see deeper what those entities hide.
435 \subsubsection pf_router router
436 As said before, <b>router</b> is used only to give some information
437 for routing algorithms. So, it does not have any attributes except :
439 <b>router</b> attributes :
440 \li <b>id (mandatory)</b>: the identifier of the router to be used
441 when referring to it.
442 \li <b>coordinates</b>: you'll have to give it if you choose the
443 vivaldi, coordinate-based routing model for the AS the host
444 belongs to. More details about it in the P2P coordinates based
447 <b>router example</b>
449 <router id="gw_dc1_horizdist"/>
452 \subsubsection pf_link link
454 Network links can represent one-hop network connections. They are
455 characterized by their id and their bandwidth. The latency is optional
456 with a default value of 0.0. For instance, we can declare a network
457 link named link1 having bandwidth of 1Gb/s and a latency of 50µs.
461 <link id="LINK1" bandwidth="125000000" latency="5E-5"/>
463 <b>Expressing sharing policy</b>
465 By default a network link is SHARED, that is if more than one flow go
466 through a link, each gets a share of the available bandwidth similar
467 to the share TCP connections offers.
469 Conversely if a link is defined as a FATPIPE, each flow going through
470 this link will get all the available bandwidth, whatever the number of
471 flows. The FATPIPE behavior allows to describe big backbones that
472 won't affect performances (except latency). Finally a link can be
473 considered as FULLDUPLEX, that means that in the simulator, 2 links
474 (one named UP and the other DOWN) will be created for each link, so as
475 the transfers from one side to the other will interact similarly as
476 TCP when ACK returning packets circulate on the other direction. More
477 discussion about it is available in <b>link_ctn</b> description.
480 <link id="SWITCH" bandwidth="125000000" latency="5E-5" sharing_policy="FATPIPE" />
483 <b>Expressing dynamicity and failures</b>
485 As for hosts, it is possible to declare links whose state, bandwidth
486 or latency change over the time. In this case, the bandwidth and
487 latency attributes are respectively replaced by the bandwidth file and
488 latency file attributes and the corresponding text files.
491 <link id="LINK1" state_file="link1.fail" bandwidth="80000000" latency=".0001" bandwidth_file="link1.bw" latency_file="link1.lat" />
494 It has to be noted that even if the syntax is the same, the semantic
495 of bandwidth and latency trace files differs from that of host
496 availability files. Those files do not express availability as a
497 fraction of the available capacity but directly in bytes per seconds
498 for the bandwidth and in seconds for the latency. This is because most
499 tools allowing to capture traces on real platforms (such as NWS)
500 express their results this way.
502 <b>Example of "link1.bw" file</b>
509 <b>Example of "link1.lat" file</b>
517 In this example, the bandwidth varies with a period of 12 seconds
518 while the latency varies with a period of 5 seconds. At the beginning
519 of simulation, the link’s bandwidth is of 80,000,000 B/s (i.e., 80
520 Mb/s). After four seconds, it drops at 40 Mb/s, and climbs back to 60
521 Mb/s after eight seconds. It keeps that way until second 12 (ie, until
522 the end of the period), point at which it loops its behavior (seconds
523 12-16 will experience 80 Mb/s, 16-20 40 Mb/s and so on). In the same
524 time, the latency values are 100µs (initial value) on the [0, 1[ time
525 interval, 1ms on [1, 2[, 10ms on [2, 3[, 1ms on [3,5[ (i.e., until the
526 end of period). It then loops back, starting at 100µs for one second.
528 <b>link</b> attributes :
529 \li <b>id (mandatory)</b>: the identifier of the link to be used when referring to it.
530 \li <b>bandwidth (mandatory)</b>: bandwidth for the link.
531 \li <b>lat </b>: latency for the link. Default is 0.0.
532 \li <b>sharing_policy</b>: sharing policy for the link.
533 \li <b>state</b>: Allow you to to set link as ON or OFF. Default is ON.
534 \li <b>bandwidth_file</b>: Allow you to use a file as input for bandwidth.
535 \li <b>latency_file</b>: Allow you to use a file as input for latency.
536 \li <b>state_file</b>: Allow you to use a file as input for states.
538 As an host, a <b>link</b> tag can also contain the <b>prop</b> tag.
542 <link id="link1" bandwidth="125000000" latency="0.000100"/>
546 \subsection pf_storage Storage
548 <b>Note : This is a prototype version that should evolve quickly, this
549 is just some doc valuable only at the time of writing this doc</b>
550 This section describes the storage management under SimGrid ; nowadays
551 it's only usable with MSG. It relies basically on linux-like concepts.
552 You also may want to have a look to its corresponding section in \ref
553 msg_file_management ; functions access are organized as a POSIX-like
556 \subsubsection pf_sto_conc Storage Main concepts
557 Basically there is 3 different entities to know :
558 \li the <b>storage_type</b>: here you define some kind of storage that
559 you will instantiate many type on your platform. Think of it like
560 a definition of throughput of a specific disk.
561 \li the <b>storage</b>: instance of a <b>storage_type</b>. Defines a
562 new storage of <b>storage_type</b>
563 \li the <b>mount</b>: says that the storage is located into this
566 the content of a storage has to be defined in a content file that
567 contains the content. The path to this file has to be passed within
568 the <b>content</b> attribute . Here is a way to generate it:
571 find /path/you/want -type f -exec ls -l {} \; 2>/dev/null > ./content.txt
574 \subsubsection pf_sto_sttp storage_type
577 <b>storage_type</b> attributes :
578 \li <b>id (mandatory)</b>: the identifier of the storage_type to be
579 used when referring to it.
580 \li <b>model (mandatory)</b>: Unused for now by the simulator (but
582 \li <b>content</b>: default value 0. The file containing the disk
583 content. (may be moved soon or later to <b>storage</b> tag.
585 The tag must contains some predefined prop, as may do some other
586 resources tags. This should moved to attributes soon or later.
587 <b>storage_type</b> mandatory <b>prop</b> :
588 \li <b>Bwrite</b>: value in B/s. Write throughput
589 \li <b>Bread</b>: value in B/s. Read throughput
590 \li <b>Bconnexion</b>: value in B/s. Connection throughput (i.e. the
591 throughput of the storage connector).
593 \subsubsection pf_sto_st storage
595 <b>storage_type</b> attributes :
596 \li <b>id (mandatory)</b>: the identifier of the storage to be used
597 when referring to it.
598 \li <b>typeId (mandatory)</b>: the identifier of the storage_type that
599 this storage belongs to.
602 \subsubsection pf_sto_mo mount
604 <b>mount</b> attributes :
605 \li <b>id (mandatory)</b>: the id of the <b>storage</b> that must be
606 mounted on that computer.
607 \li <b>name (mandatory)</b>: the name that will be the logical
608 reference to this disk (the mount point).
610 \subsubsection pf_sto_mst mstorage
611 <b>Note : unused for now</b>
612 <b>mstorage</b> attributes :
613 \li <b>typeId (mandatory)</b>: the id of the <b>storage</b> that must
614 be mounted on that computer.
615 \li <b>name (mandatory)</b>: the name that will be the logical
616 reference to this disk (the mount point).
618 \section pf_routing Routing
620 In order to run fast, it has been chosen to use static routing within
621 SimGrid. By static, it means that it is calculated once (or almost),
622 and will not change during execution. We chose to do that because it
623 is rare to have a real deficience of a resource ; most of the time, a
624 communication fails because the links are too overloaded, and so your
625 connection stops before the time out, or because the computer at the
626 other end is not answering.
628 We also chose to use shortests paths algorithms in order to emulate
629 routing. Doing so is consistent with the reality: RIP, OSPF, BGP are
630 all calculating shortest paths. They have some convergence time, but
631 at the end, so when the platform is stable (and this should be the
632 moment you want to simulate something using SimGrid) your packets will
633 follow the shortest paths.
635 \subsection pf_rm Routing models
637 Within each AS, you have to define a routing model to use. You have
638 basically 3 main kind of routing models :
640 \li Shortest-path based models: you let SimGrid calculates shortest
641 paths and manage it. Behaves more or less as most real life
643 \li Manually-entered route models: you'll have to define all routes
644 manually by yourself into the platform description file.
645 Consistent with some manually managed real life routing.
646 \li Simple/fast models: those models offers fast, low memory routing
647 algorithms. You should consider to use it if you can make some
648 assumptions about your AS. Routing in this case is more or less
651 \subsubsection pf_raf The router affair
653 Expressing routers becomes mandatory when using shortest-path based
654 models or when using ns-3 or the bindings to the GTNetS packet-level
655 simulator instead of the native analytical network model implemented
658 For graph-based shortest path algorithms, routers are mandatory,
659 because both algorithms need a graph, and so we need to have source
660 and destination for each edge.
662 Routers are naturally an important concept in GTNetS or ns-3 since the
663 way they run the packet routing algorithms is actually simulated.
664 Instead, the SimGrid’s analytical models aggregate the routing time
665 with the transfer time. Rebuilding a graph representation only from
666 the route information turns to be a very difficult task, because of
667 the missing information about how routes intersect. That is why we
668 introduced a \<router\> tag, which is simply used to express these
669 intersection points. The only attribute accepted by this tag an id. It
670 is important to understand that the \<router\> tag is only used to
671 provide topological information.
673 To express those topological information, some <b>route</b> have to be
674 defined saying which link is between which routers. Description or the
675 route syntax is given below, as well as example for the different
678 \subsubsection pf_rm_sh Shortest-path based models
680 Here is the complete list of such models, that computes routes using
681 classic shortest-paths algorithms. How to choose the best suited
682 algorithm is discussed later in the section devoted to it.
684 \li <b>Floyd</b>: Floyd routing data. Pre-calculates all routes once.
685 \li <b>Dijkstra</b>: Dijkstra routing data ,calculating routes when
687 \li <b>DijkstraCache</b>: Dijkstra routing data. Handle some cache for
688 already calculated routes.
690 All those shortest-path models are instanciated the same way. Here are
695 <AS id="AS0" routing="Floyd">
697 <cluster id="my_cluster_1" prefix="c-" suffix=""
698 radical="0-1" power="1000000000" bw="125000000" lat="5E-5"
699 router_id="router1"/>
701 <AS id="AS1" routing="none">
702 <host id="host1" power="1000000000"/>
705 <link id="link1" bandwidth="100000" latency="0.01"/>
707 <ASroute src="my_cluster_1" dst="AS1"
710 <link_ctn id="link1"/>
716 ASroute given at the end gives a topological information: link1 is
717 between router1 and host1.
721 <AS id="AS_2" routing="Dijsktra">
722 <host id="AS_2_host1" power="1000000000"/>
723 <host id="AS_2_host2" power="1000000000"/>
724 <host id="AS_2_host3" power="1000000000"/>
725 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
726 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
727 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
728 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
729 <router id="central_router"/>
730 <router id="AS_2_gateway"/>
731 <!-- routes providing topological information -->
732 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
733 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
734 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
735 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
739 DijsktraCache example :
741 <AS id="AS_2" routing="DijsktraCache">
742 <host id="AS_2_host1" power="1000000000"/>
744 (platform unchanged compared to upper example)
747 \subsubsection pf_rm_me Manually-entered route models
749 \li <b>Full</b>: You have to enter all necessary routes manually
750 \li <b>RuleBased</b>: Rule-Based routing data; same as Full except you
751 can use regexp to express route. As SimGrid has to evaluate the
752 regexp, it's slower than Full, but requires less memory. Regexp
753 syntax is similar as <a href="http://www.pcre.org">pcre</a> ones,
754 as this is the lib SimGrid use to do so.
759 <AS id="AS0" routing="Full">
760 <host id="host1" power="1000000000"/>
761 <host id="host2" power="1000000000"/>
762 <link id="link1" bandwidth="125000000" latency="0.000100"/>
763 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
769 <AS id="AS_orsay" routing="RuleBased" >
770 <cluster id="AS_gdx" prefix="gdx-" suffix=".orsay.grid5000.fr"
771 radical="1-310" power="4.7153E9" bw="1.25E8" lat="1.0E-4"
772 bb_bw="1.25E9" bb_lat="1.0E-4"></cluster>
773 <link id="link_gdx" bandwidth="1.25E9" latency="1.0E-4"/>
775 <cluster id="AS_netgdx" prefix="netgdx-" suffix=".orsay.grid5000.fr"
776 radical="1-30" power="4.7144E9" bw="1.25E8" lat="1.0E-4"
777 bb_bw="1.25E9" bb_lat="1.0E-4"></cluster>
778 <link id="link_netgdx" bandwidth="1.25E9" latency="1.0E-4"/>
780 <AS id="gw_AS_orsay" routing="Full">
781 <router id="gw_orsay"/>
783 <link id="link_gw_orsay" bandwidth="1.25E9" latency="1.0E-4"/>
785 <ASroute src="^AS_(.*)$" dst="^AS_(.*)$"
786 gw_src="$1src-AS_$1src_router.orsay.grid5000.fr"
787 gw_dst="$1dst-AS_$1dst_router.orsay.grid5000.fr"
789 <link_ctn id="link_$1src"/>
790 <link_ctn id="link_$1dst"/>
793 <ASroute src="^AS_(.*)$" dst="^gw_AS_(.*)$"
794 gw_src="$1src-AS_$1src_router.orsay.grid5000.fr"
797 <link_ctn id="link_$1src"/>
800 <ASroute src="^gw_AS_(.*)$" dst="^AS_(.*)$"
802 gw_dst="$1dst-AS_$1dst_router.orsay.grid5000.fr"
804 <link_ctn id="link_$1dst"/>
810 The example upper contains $1src and $1dst. It's simply a reference to
811 string matching regexp enclosed by "()" within respectively <b>src</b>
812 and <b>dst</b> attributes. If they were more than 1 "()", then you
813 could referer to it as $2src, $3src and so on.
815 \subsubsection pf_rm_sf Simple/fast models
817 \li <b>none</b>: No routing (Unless you know what you are doing, avoid
818 using this mode in combination with a non Constant network model).
821 <AS id="exitAS" routing="none">
822 <router id="exit_gateway"/>
825 \li <b>Vivaldi</b>: Vivaldi routing, so when you want to use
826 coordinates. See the corresponding section P2P below for details.
827 \li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
828 not be used, except internally.
830 \subsection ps_dec Defining routes
832 The principle of route definition is the same for the 4 available tags
833 for doing it. Those for tags are:
835 \li <b>route</b>: to define route between host/router
836 \li <b>ASroute</b>: to define route between AS
837 \li <b>bypassRoute</b>: to bypass normal routes as calculated by the
838 network model between host/router
839 \li <b>bypassASroute</b>: same as bypassRoute, but for AS
841 Basically all those tags will contain an (ordered) list of references
842 to link that compose the route you want to define.
844 Consider the example below:
847 <route src="Alice" dst="Bob">
848 <link_ctn id="link1"/>
849 <link_ctn id="link2"/>
850 <link_ctn id="link3"/>
854 The route here fom host Alice to Bob will be first link1, then link2,
855 and finally link3. What about the reverse route ? <b>route</b> and
856 <b>ASroute</b> have an optional attribute <b>symmetrical</b>, that can
857 be either YES or NO. YES means that the reverse route is the same
858 route in the inverse order, and is setted to YES by default. Note that
859 this is not the case for bypass*Route, as it is more probable that you
860 want to bypass only one default route.
862 For an ASroute, things are just sligthly more complicated, as you have
863 to give the id of the gateway which is inside the AS you're talking
864 about you want to access ... So it looks like this :
868 <ASroute src="AS1" dst="AS2"
869 gw_src="router1" gw_dst="router2">
870 <link_ctn id="link1"/>
874 gw == gateway, so when any message are trying to go from AS1 to AS2,
875 it means that it must pass through router1 to get out of the AS, then
876 pass through link1, and get into AS2 by being received by router2.
877 router1 must belong to AS1 and router2 must belong to AS2.
879 \subsubsection pf_linkctn link_ctn
881 a <b>link_ctn</b> is the tag that is used in order to reference a
882 <b>link</b> in a route. Its id is the link id it refers to.
884 <b>link_ctn</b> attributes :
885 \li <b>id (mandatory)</b>: Id of the link this tag refers to
886 \li <b>direction</b>: if the link referenced by <b>id</b> has been
887 declared as FULLDUPLEX, this is used to indicate in which
888 direction the route you're defining is going through this link.
889 Possible values "UP" or "DOWN".
891 \subsubsection pf_asro ASroute
893 ASroute tag purpose is to let people write manually their routes
894 between AS. It's usefull when you're in Full or Rule-based model.
896 <b>ASroute</b> attributes :
897 \li <b>src (mandatory)</b>: the source AS id.
898 \li <b>dst (mandatory)</b>: the destination AS id.
899 \li <b>gw_src (mandatory)</b>: the gateway to be used within the AS.
900 Can be any <b>host</b> or \b router defined into the \b src AS or
901 into one of the AS it includes.
902 \li <b>gw_dst (mandatory)</b>: the gateway to be used within the AS.
903 Can be any <b>host</b> or \b router defined into the \b dst AS or
904 into one of the AS it includes.
905 \li <b>symmetrical</b>: if the route is symmetric, the reverse route
906 will be the opposite of the one defined. Can be either YES or NO,
909 <b>Example of ASroute with RuleBased</b>
911 <ASroute src="^gw_AS_(.*)$" dst="^AS_(.*)$"
913 gw_dst="$1dst-AS_$1dst_router.orsay.grid5000.fr"
915 <link_ctn id="link_$1dst"/>
918 <b>Example of ASroute with Full</b>
920 <AS id="AS0" routing="Full">
921 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
922 radical="0-149" power="1000000000" bw="125000000" lat="5E-5"
923 bb_bw="2250000000" bb_lat="5E-4"/>
925 <cluster id="my_cluster_2" prefix="c-" suffix=".me"
926 radical="150-299" power="1000000000" bw="125000000" lat="5E-5"
927 bb_bw="2250000000" bb_lat="5E-4"/>
929 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
931 <ASroute src="my_cluster_1" dst="my_cluster_2"
932 gw_src="c-my_cluster_1_router.me"
933 gw_dst="c-my_cluster_2_router.me">
934 <link_ctn id="backbone"/>
936 <ASroute src="my_cluster_2" dst="my_cluster_1"
937 gw_src="c-my_cluster_2_router.me"
938 gw_dst="c-my_cluster_1_router.me">
939 <link_ctn id="backbone"/>
944 \subsubsection pf_ro route
945 The principle is the same as ASroute : <b>route</b> contains list of
946 links that are in the path between src and dst, except that it is for
947 routes between a src that can be either <b>host</b> or \b router and a
948 dst that can be either <b>host</b> or \b router. Usefull for Full and
949 RuleBased, as well as for the shortest-paths based models, where you
950 have to give topological informations.
953 <b>route</b> attributes :
954 \li <b>src (mandatory)</b>: the source id.
955 \li <b>dst (mandatory)</b>: the destination id.
956 \li <b>symmetrical</b>: if the route is symmetric, the reverse route
957 will be the opposite of the one defined. Can be either YES or NO,
960 <b>route example in Full</b>
962 <route src="Tremblay" dst="Bourassa">
963 <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"/>
967 <b>route example in a shortest-path model</b>
969 <route src="Tremblay" dst="Bourassa">
973 Note that when using route to give topological information, you have
974 to give routes with one link only in it, as SimGrid needs to know
975 which host are at the end of the link.
977 \subsubsection pf_byro bypassASroute
979 <b>Note : bypassASroute and bypassRoute are under rewriting to perform
980 better ; so you may not use it yet</b> As said before, once you choose
981 a model, it (if so) calculates routes for you. But maybe you want to
982 define some of your routes, which will be specific. You may also want
983 to bypass some routes defined in lower level AS at an upper stage :
984 <b>bypassASroute</b> is the tag you're looking for. It allows to
985 bypass routes defined between already defined between AS (if you want
986 to bypass route for a specific host, you should just use byPassRoute).
987 The principle is the same as ASroute : <b>bypassASroute</b> contains
988 list of links that are in the path between src and dst.
990 <b>bypassASroute</b> attributes :
991 \li <b>src (mandatory)</b>: the source AS id.
992 \li <b>dst (mandatory)</b>: the destination AS id.
993 \li <b>gw_src (mandatory)</b>: the gateway to be used within the AS.
994 Can be any <b>host</b> or \b router defined into the \b src AS or
995 into one of the AS it includes.
996 \li <b>gw_dst (mandatory)</b>: the gateway to be used within the AS.
997 Can be any <b>host</b> or \b router defined into the \b dst AS or
998 into one of the AS it includes.
999 \li <b>symmetrical</b>: if the route is symmetric, the reverse route
1000 will be the opposite of the one defined. Can be either YES or NO,
1003 <b>bypassASroute Example</b>
1005 <bypassASRoute src="my_cluster_1" dst="my_cluster_2"
1006 gw_src="my_cluster_1_router"
1007 gw_dst="my_cluster_2_router">
1008 <link_ctn id="link_tmp"/>
1012 \subsubsection pf_byro bypassRoute
1013 <b>Note : bypassASRoute and bypassRoute are under rewriting to perform
1014 better ; so you may not use it yet</b> As said before, once you choose
1015 a model, it (if so) calculates routes for you. But maybe you want to
1016 define some of your routes, which will be specific. You may also want
1017 to bypass some routes defined in lower level AS at an upper stage :
1018 <b>bypassRoute</b> is the tag you're looking for. It allows to bypass
1019 routes defined between <b>host/router</b>. The principle is the same
1020 as route : <b>bypassRoute</b> contains list of links references of
1021 links that are in the path between src and dst.
1023 <b>bypassRoute</b> attributes :
1024 \li <b>src (mandatory)</b>: the source AS id.
1025 \li <b>dst (mandatory)</b>: the destination AS id.
1026 \li <b>symmetrical</b>: if the route is symmetric, the reverse route
1027 will be the opposite of the one defined. Can be either YES or NO,
1030 <b>bypassRoute Example</b>
1032 <b>bypassRoute Example</b>
1034 <bypassRoute src="host_1" dst="host_2">
1035 <link_ctn id="link_tmp"/>
1040 \subsection pb_baroex Basic Routing Example
1042 Let's say you have an AS named AS_Big that contains two other AS, AS_1
1043 and AS_2. If you want to make an host (h1) from AS_1 with another one
1044 (h2) from AS_2 then you'll have to proceed as follow:
1045 \li First, you have to ensure that a route is defined from h1 to the
1046 AS_1's exit gateway and from h2 to AS_2's exit gateway.
1047 \li Then, you'll have to define a route between AS_1 to AS_2. As those
1048 AS are both resources belonging to AS_Big, then it has to be done
1049 at AS_big level. To define such a route, you have to give the
1050 source AS (AS_1), the destination AS (AS_2), and their respective
1051 gateway (as the route is effectively defined between those two
1052 entry/exit points). Elements of this route can only be elements
1053 belonging to AS_Big, so links and routers in this route should be
1054 defined inside AS_Big. If you choose some shortest-path model,
1055 this route will be computed automatically.
1057 As said before, there are mainly 2 tags for routing :
1058 \li <b>ASroute</b>: to define routes between two <b>AS</b>
1059 \li <b>route</b>: to define routes between two <b>host/router</b>
1061 As we are dealing with routes between AS, it means that those we'll
1062 have some definition at AS_Big level. Let consider AS_1 contains 1
1063 host, 1 link and one router and AS_2 3 hosts, 4 links and one router.
1064 There will be a central router, and a cross-like topology. At the end
1065 of the crosses arms, you'll find the 3 hosts and the router that will
1066 act as a gateway. We have to define routes inside those two AS. Let
1067 say that AS_1 contains full routes, and AS_2 contains some Floyd
1068 routing (as we don't want to bother with defining all routes). As
1069 we're using some shortest path algorithms to route into AS_2, we'll
1070 then have to define some <b>route</b> to gives some topological
1071 information to SimGrid. Here is a file doing it all :
1074 <AS id="AS_Big" routing="Dijsktra">
1075 <AS id="AS_1" routing="Full">
1076 <host id="AS_1_host1" power="1000000000"/>
1077 <link id="AS_1_link" bandwidth="1250000000" latency="5E-4"/>
1078 <router id="AS_1_gateway"/>
1079 <route src="AS_1_host1" dst="AS_1_gateway">
1080 <link_ctn id="AS_1_link"/>
1083 <AS id="AS_2" routing="Floyd">
1084 <host id="AS_2_host1" power="1000000000"/>
1085 <host id="AS_2_host2" power="1000000000"/>
1086 <host id="AS_2_host3" power="1000000000"/>
1087 <link id="AS_2_link1" bandwidth="1250000000" latency="5E-4"/>
1088 <link id="AS_2_link2" bandwidth="1250000000" latency="5E-4"/>
1089 <link id="AS_2_link3" bandwidth="1250000000" latency="5E-4"/>
1090 <link id="AS_2_link4" bandwidth="1250000000" latency="5E-4"/>
1091 <router id="central_router"/>
1092 <router id="AS_2_gateway"/>
1093 <!-- routes providing topological information -->
1094 <route src="central_router" dst="AS_2_host1"><link_ctn id="AS_2_link1"/></route>
1095 <route src="central_router" dst="AS_2_host2"><link_ctn id="AS_2_link2"/></route>
1096 <route src="central_router" dst="AS_2_host3"><link_ctn id="AS_2_link3"/></route>
1097 <route src="central_router" dst="AS_2_gateway"><link_ctn id="AS_2_link4"/></route>
1099 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1101 <ASroute src="AS_1" dst="AS_2"
1102 gw_src="AS_1_gateway"
1103 gw_dst="AS_2_gateway">
1104 <link_ctn id="backbone"/>
1109 \section pf_other_tags Tags not (directly) describing the platform
1111 There are 3 tags, that you can use inside a \<platform\> tag that are
1112 not describing the platform:
1113 \li random: it allows you to define random generators you want to use
1114 for your simulation.
1115 \li config: it allows you to pass some configuration stuff like, for
1116 example, the network model and so on. It follows the
1117 \li include: simply allows you to include another file into the
1120 \subsection pf_conf config
1121 <b>config</b> attributes :
1122 \li <b>id (mandatory)</b>: the identifier of the config to be used
1123 when referring to it.
1126 <b>config</b> tag only purpose is to include <b>prop</b> tags. Valid
1127 id are basically the same as the list of possible parameters you can
1128 use by command line, except that "/" are used for namespace
1129 definition. See the \ref options config and options page for more
1133 <b>config example</b>
1135 <?xml version='1.0'?>
1136 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1137 <platform version="3">
1138 <config id="General">
1139 <prop id="maxmin/precision" value="0.000010"></prop>
1140 <prop id="cpu/optim" value="TI"></prop>
1141 <prop id="workstation/model" value="compound"></prop>
1142 <prop id="network/model" value="SMPI"></prop>
1143 <prop id="path" value="~/"></prop>
1144 <prop id="smpi/bw_factor" value="65472:0.940694;15424:0.697866;9376:0.58729"></prop>
1147 <AS id="AS0" routing="Full">
1152 \subsection pf_rand random
1153 Not yet in use, and possibly subject to huge modifications.
1155 \subsection pf_incl include
1156 <b>include</b> tag allows to import into a file platform parts located
1157 in another file. This is done with the intention to help people
1158 combine their different AS and provide new platforms. Those files
1159 should contains XML part that contains either
1160 <b>include,cluster,peer,AS,trace,trace_connect</b> tags.
1162 <b>include</b> attributes :
1163 \li <b>file (mandatory)</b>: filename of the file to include. Possible
1164 values: absolute or relative path, syntax similar to the one in
1167 <b>Note</b>: due to some obscure technical reasons, you have to open
1168 and close tag in order to let it work.
1169 <b>include Example</b>
1171 <?xml version='1.0'?>
1172 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1173 <platform version="3">
1174 <AS id="main" routing="Full">
1175 <include file="clusterA.xml"></include>
1176 <include file="clusterB.xml"></include>
1181 \subsection pf_tra trace and trace_connect
1182 Both tags are an alternate way to passe availability, state, and so on
1183 files to entity. Instead of refering to the file directly in the host,
1184 link, or cluster tag, you proceed by defining a trace with an id
1185 corresponding to a file, later an host/link/cluster, and finally using
1186 trace_connect you say that the file trace must be used by the entity.
1187 Get it ? Let's have a look at an example :
1190 <AS id="AS0" routing="Full">
1191 <host id="bob" power="1000000000"/>
1193 <trace id="myTrace" file="bob.trace" periodicity="1.0"/>
1194 <trace_connect trace="myTrace" element="bob" kind="POWER"/>
1197 All constraints you have is that <b>trace_connect</b> is after
1198 <b>trace</b> and <b>host</b> definitions.
1201 <b>trace</b> attributes :
1202 \li <b>id (mandatory)</b>: the identifier of the trace to be used when
1204 \li <b>file</b>: filename of the file to include. Possible values :
1205 absolute or relative path, syntax similar to the one in use on
1206 your system. If ommited, the system expects that you provide the
1207 trace values inside the trace tags (see below).
1208 \li <b>trace periodicity (mandatory)</b>: trace periodicity, same
1209 definition as in hosts (see upper for details).
1211 Here is an example of trace when no file name is provided:
1214 <trace id="myTrace" periodicity="1.0">
1221 <b>trace_connect</b> attributes :
1222 \li <b>kind</b>: the type of trace, possible values
1223 <b>HOST_AVAIL|POWER|LINK_AVAIL|BANDWIDTH|LATENCY,</b> default:
1225 \li <b>trace (mandatory)</b>: the identifier of the trace referenced.
1226 \li <b>element (mandatory)</b>: the identifier of the entity referenced.
1230 \section pf_hints Hints and tips, or how to write a platform efficiently
1232 Now you should know at least the syntax dans be able to create a
1233 platform. However, after having ourselves wrote some platforms, there
1234 are some best practices you should pay attention to in order to
1235 produce good platform and some choices you can make in order to have
1236 faster simulations. Here's some hints and tips, then.
1238 \subsection pf_as_h AS Hierarchy
1239 The AS design allows SimGrid to go fast, because computing route is
1240 done only for the set of resources defined in this AS. If you're using
1241 only a big AS containing all resource with no AS into it and you're
1242 using Full model, then ... you'll loose all interest into it. On the
1243 other hand, designing a binary tree of AS with, at the lower level,
1244 only one host, then you'll also loose all the good AS hierarchy can
1245 give you. Remind you should always be "reasonable" in your platform
1246 definition when choosing the hierarchy. A good choice if you try to
1247 describe a real life platform is to follow the AS described in
1248 reality, since this kind og trade-off works well for real life
1251 \subsection pf_exit_as Exit AS: why and how
1252 Users that have looked at some of our platforms may have notice a
1253 non-intuitive schema ... Something like that :
1257 <AS id="AS_4" routing="Full">
1258 <AS id="exitAS_4" routing="Full">
1259 <router id="router_4"/>
1261 <cluster id="cl_4_1" prefix="c_4_1-" suffix="" radical="1-20" power="1000000000" bw="125000000" lat="5E-5" bb_bw="2250000000" bb_lat="5E-4"/>
1262 <cluster id="cl_4_2" prefix="c_4_2-" suffix="" radical="1-20" power="1000000000" bw="125000000" lat="5E-5" bb_bw="2250000000" bb_lat="5E-4"/>
1263 <link id="4_1" bandwidth="2250000000" latency="5E-5"/>
1264 <link id="4_2" bandwidth="2250000000" latency="5E-5"/>
1265 <link id="bb_4" bandwidth="2250000000" latency="5E-4"/>
1266 <ASroute src="cl_4_1"
1268 gw_src="c_4_1-cl_4_1_router"
1269 gw_dst="c_4_2-cl_4_2_router"
1271 <link_ctn id="4_1"/>
1272 <link_ctn id="bb_4"/>
1273 <link_ctn id="4_2"/>
1275 <ASroute src="cl_4_1"
1277 gw_src="c_4_1-cl_4_1_router"
1280 <link_ctn id="4_1"/>
1281 <link_ctn id="bb_4"/>
1283 <ASroute src="cl_4_2"
1285 gw_src="c_4_2-cl_4_2_router"
1288 <link_ctn id="4_2"/>
1289 <link_ctn id="bb_4"/>
1294 In the AS_4, you have an exitAS_4 defined, containing only one router,
1295 and routes defined to that AS from all other AS (as cluster is only a
1296 shortcut for an AS, see cluster description for details). If there was
1297 an upper AS, it would define routes to and from AS_4 with the gateway
1298 router_4. It's just because, as we did not allowed (for performances
1299 issues) to have routes from an AS to a single host/router, you have to
1300 enclose your gateway, when you have AS included in your AS, within an
1301 AS to define routes to it.
1303 \subsection pf_P2P_tags P2P or how to use coordinates
1304 SimGrid allows you to use some coordinated-based system, like vivaldi,
1305 to describe a platform. The main concept is that you have some peers
1306 that are located somewhere: this is the function of the
1307 <b>coordinates</b> of the \<peer\> or \<host\> tag. There's nothing
1308 complicated in using it, here is an example of it:
1311 <?xml version='1.0'?>
1312 <!DOCTYPE platform SYSTEM "http://simgrid.gforge.inria.fr/simgrid.dtd">
1313 <platform version="3">
1315 <config id="General">
1316 <prop id="network/coordinates" value="yes"></prop>
1318 <AS id="AS0" routing="Vivaldi">
1319 <host id="100030591" coordinates="25.5 9.4 1.4" power="1500000000.0" />
1320 <host id="100036570" coordinates="-12.7 -9.9 2.1" power="730000000.0" />
1322 <host id="100429957" coordinates="17.5 6.7 18.8" power="830000000.0" />
1327 Coordinates are then used to calculate latency between two hosts by
1328 calculating the euclidian distance between the two hosts coordinates.
1329 The results express the latency in ms.
1331 \subsection pf_wisely Choosing wisely the routing model to use
1334 Choosing wisely the routing model to use can significantly fasten your
1335 simulation/save your time when writing the platform/save tremendeous
1336 disk space. Here is the list of available model and their
1337 characteristics (lookup : time to resolve a route):
1339 \li <b>Full</b>: Full routing data (fast, large memory requirements,
1341 \li <b>Floyd</b>: Floyd routing data (slow initialization, fast
1342 lookup, lesser memory requirements, shortest path routing only).
1343 Calculates all routes at once at the beginning.
1344 \li <b>Dijkstra</b>: Dijkstra routing data (fast initialization, slow
1345 lookup, small memory requirements, shortest path routing only).
1346 Calculates a route when necessary.
1347 \li <b>DijkstraCache</b>: Dijkstra routing data (fast initialization,
1348 fast lookup, small memory requirements, shortest path routing
1349 only). Same as Dijkstra, except it handles a cache for latest used
1351 \li <b>none</b>: No routing (usable with Constant network only).
1352 Defines that there is no routes, so if you try to determine a
1353 route without constant network within this AS, SimGrid will raie
1355 \li <b>RuleBased</b>: Rule-Based routing data (fast initialisation,
1356 relatively slow lookup, moderate memory requirements, fully
1357 expressive): uses regexp to define routes;
1358 \li <b>Vivaldi</b>: Vivaldi routing, so when you want to use coordinates
1359 \li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
1362 \subsection pf_switch Hey, I want to describe a switch but there is no switch tag !
1364 Actually we did not include swith tag, ok. But when you're trying to
1365 simulate a switch, the only major impact it has when you're using
1366 fluid model (and SimGrid uses fluid model unless you activate GTNetS,
1367 ns-3, or constant network mode) is the impact of the upper limit of
1368 the switch motherboard speed that will eventually be reached if you're
1369 using intensively your switch. So, the switch impact is similar to a
1370 link one. That's why we are used to describe a switch using a link tag
1371 (as a link is not an edge by a hyperedge, you can connect more than 2
1374 \subsection pf_platform_multipath How to express multipath routing in platform files?
1376 It is unfortunately impossible to express the fact that there is more
1377 than one routing path between two given hosts. Let's consider the
1378 following platform file:
1381 <route src="A" dst="B">
1384 <route src="B" dst="C">
1387 <route src="A" dst="C">
1392 Although it is perfectly valid, it does not mean that data traveling
1393 from A to C can either go directly (using link 3) or through B (using
1394 links 1 and 2). It simply means that the routing on the graph is not
1395 trivial, and that data do not following the shortest path in number of
1396 hops on this graph. Another way to say it is that there is no implicit
1397 in these routing descriptions. The system will only use the routes you
1398 declare (such as <route src="A" dst="C"><link_ctn
1399 id="3"/></route>), without trying to build new routes by aggregating
1402 You are also free to declare platform where the routing is not
1403 symmetric. For example, add the following to the previous file:
1406 <route src="C" dst="A">
1412 This makes sure that data from C to A go through B where data from A
1413 to C go directly. Don't worry about realism of such settings since
1414 we've seen ways more weird situation in real settings (in fact, that's
1415 the realism of very regular platforms which is questionable, but
1416 that's another story).
1418 \section pf_flexml_bypassing Bypassing the XML parser with your own C functions
1419 <b>NOTE THAT THIS DOCUMENTATION, WHILE STILL WORKING, IS STRONGLY DEPRECATED</b>
1421 So you want to bypass the XML files parser, uh? Maybe doing some parameter
1422 sweep experiments on your simulations or so? This is possible, and
1423 it's not even really difficult (well. Such a brutal idea could be
1424 harder to implement). Here is how it goes.
1426 For this, you have to first remember that the XML parsing in SimGrid is done
1427 using a tool called FleXML. Given a DTD, this gives a flex-based parser. If
1428 you want to bypass the parser, you need to provide some code mimicking what
1429 it does and replacing it in its interactions with the SURF code. So, let's
1430 have a look at these interactions.
1432 FleXML parser are close to classical SAX parsers. It means that a
1433 well-formed SimGrid platform XML file might result in the following
1436 - start "platform_description" with attribute version="2"
1437 - start "host" with attributes id="host1" power="1.0"
1439 - start "host" with attributes id="host2" power="2.0"
1441 - start "link" with ...
1443 - start "route" with ...
1444 - start "link_ctn" with ...
1447 - end "platform_description"
1449 The communication from the parser to the SURF code uses two means:
1450 Attributes get copied into some global variables, and a surf-provided
1451 function gets called by the parser for each event. For example, the event
1452 - start "host" with attributes id="host1" power="1.0"
1454 let the parser do something roughly equivalent to:
1456 strcpy(A_host_id,"host1");
1461 In SURF, we attach callbacks to the different events by initializing the
1462 pointer functions to some the right surf functions. Since there can be
1463 more than one callback attached to the same event (if more than one
1464 model is in use, for example), they are stored in a dynar. Example in
1465 workstation_ptask_L07.c:
1467 /* Adding callback functions */
1468 surf_parse_reset_parser();
1469 surfxml_add_callback(STag_surfxml_host_cb_list, &parse_cpu_init);
1470 surfxml_add_callback(STag_surfxml_prop_cb_list, &parse_properties);
1471 surfxml_add_callback(STag_surfxml_link_cb_list, &parse_link_init);
1472 surfxml_add_callback(STag_surfxml_route_cb_list, &parse_route_set_endpoints);
1473 surfxml_add_callback(ETag_surfxml_link_c_ctn_cb_list, &parse_route_elem);
1474 surfxml_add_callback(ETag_surfxml_route_cb_list, &parse_route_set_route);
1476 /* Parse the file */
1477 surf_parse_open(file);
1478 xbt_assert(!surf_parse(), "Parse error in %s", file);
1482 So, to bypass the FleXML parser, you need to write your own version of the
1483 surf_parse function, which should do the following:
1484 - Fill the A_<tag>_<attribute> variables with the wanted values
1485 - Call the corresponding STag_<tag>_fun function to simulate tag start
1486 - Call the corresponding ETag_<tag>_fun function to simulate tag end
1487 - (do the same for the next set of values, and loop)
1489 Then, tell SimGrid that you want to use your own "parser" instead of the stock one:
1491 surf_parse = surf_parse_bypass_environment;
1492 MSG_create_environment(NULL);
1493 surf_parse = surf_parse_bypass_application;
1494 MSG_launch_application(NULL);
1497 A set of macros are provided at the end of
1498 include/surf/surfxml_parse.h to ease the writing of the bypass
1499 functions. An example of this trick is distributed in the file
1500 examples/msg/masterslave/masterslave_bypass.c