1 /*! @page platform Describing the virtual platform
5 As usual, SimGrid is a versatile framework, and you should find the
6 way of describing your platform that best fits your experimental
9 @section pf_first_example First Platform Example
11 Here is a very simple platform file, containing 3 resources (two hosts
12 and one link), and explicitly giving the route between the hosts.
16 <!DOCTYPE platform SYSTEM "https://simgrid.org/simgrid.dtd">
17 <platform version="4.1">
18 <zone id="first zone" routing="Full">
19 <!-- the resources -->
20 <host id="host1" speed="1Mf"/>
21 <host id="host2" speed="2Mf"/>
22 <link id="link1" bandwidth="125MBps" latency="100us"/>
23 <!-- the routing: specify how the hosts are interconnected -->
24 <route src="host1" dst="host2">
25 <link_ctn id="link1"/>
32 Then, every resource (specified with @ref pf_tag_host, @ref
33 pf_tag_link or others) must be located within a given **networking
34 zone**. Each zone is in charge of the routing between its
35 resources. It means that when an host wants to communicate with
36 another host of the same zone, it is the zone's duty to find the list
37 of links that are involved in the communication. Here, since the @ref
38 pf_tag_zone tag has **Full** as a **routing attribute**, all routes
39 must be explicitely given using the @ref pf_tag_route and @ref
40 pf_tag_linkctn tags (this @ref pf_rm "routing model" is both simple
41 and inefficient :) It is OK to not specify the route between two
42 hosts, as long as the processes located on them never try to
45 A zone can contain several zones itself, leading to a hierarchical
46 decomposition of the platform. This can be more efficient (as the
47 inter-zone routing gets factorized with @ref pf_tag_zoneroute), and
48 allows to have more than one routing model in your platform. For
49 example, you could have a coordinate-based routing for the WAN parts
50 of your platforms, a full routing within each datacenter, and a highly
51 optimized routing within each cluster of the datacenter. In this
52 case, determining the route between two given hosts gets @ref
53 routing_basics "somewhat more complex" but SimGrid still computes
54 these routes for you in a time- and space-efficient manner.
55 Here is an illustration of these concepts:
57 
59 Circles represent processing units and squares represent network
60 routers. Bold lines represent communication links. The zone "AS2"
61 models the core of a national network interconnecting a small flat
62 cluster (AS4) and a larger hierarchical cluster (AS5), a subset of a
63 LAN (AS6), and a set of peers scattered around the world (AS7).
65 @section pf_res Resource description
67 @subsection pf_res_computing Computing Resources
69 @subsubsection pf_tag_host <host>
71 An host is the computing resource on which an actor can execute.
73 Attribute | Values | Description
74 ----------------- | -------------------------------------- | -----------
75 id | String (mandatory) | The identifier of the host. facilitates referring to this AS.
76 speed | double (mandatory) | Computational power of every core of this host in FLOPS (must be positive)
77 core | int (defaults to 1) | Number of cores (see @ref howto_multicore)
78 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.
79 state_file | File name (optional) | File to use as a state profile (see @ref howto_churn)
80 coordinates | String (mandatory when using Vivaldi routing) | The coordinates of this host (see @ref pf_P2P_tags).
81 pstate | Double (Defaults to 0) | FIXME: Not yet documented.
83 #### Included tags ####
85 - @ref pf_tag_mount Specifies the storages mounted on that host
86 - @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().
91 <host id="host1" speed="1000000000"/>
92 <host id="host2" speed="1000000000">
93 <prop id="color" value="blue"/>
94 <prop id="rendershape" value="square"/>
98 @anchor pf_host_dynamism
99 ### Expressing dynamism ###
101 SimGrid provides mechanisms to change a hosts' availability over
102 time, using the ``availability_file`` attribute to the ``@<host@>`` tag
103 and a separate text file whose syntax is exemplified below.
105 #### Adding a trace file ####
108 <platform version="4">
109 <host id="bob" speed="500Gf" availability_file="bob.trace" />
113 #### Example of "bob.trace" file ####
122 Let us begin to explain this example by looking at line 2. (Line 1 will become clear soon).
123 The first column describes points in time, in this case, time 0. The second column
124 describes the relative amount of power this host is able to deliver (relative
125 to the maximum performance specified in the ``@<host@>`` tag). (Clearly, the
126 second column needs to contain values that are not smaller than 0 and not larger than 1).
127 In this example, our host will deliver 500 Mflop/s at time 0, as 500 Mflop/s is the
128 maximum performance of this host. At time 11.0, it will
129 deliver half of its maximum performance, i.e., 250 Mflop/s until time 20.0 when it will
130 will start delivering 80@% of its power. In this example, this amounts to 400 Mflop/s.
132 Since the periodicity in line 1 was set to be 1.0, i.e., 1 timestep, this host will
133 continue to provide 500 Mflop/s from time 21. From time 32 it will provide 250 MFlop/s and so on.
135 @subsubsection pf_tag_cluster <cluster>
137 ``<cluster />`` represents a machine-cluster. It is most commonly used
138 when one wants to define many hosts and a network quickly. Technically,
139 ``cluster`` is a meta-tag: <b>from the inner SimGrid point of
140 view, a cluster is a network zone where some optimized routing is defined</b>.
141 The default inner organization of the cluster is as follow:
147 ____________|__________|_____________ backbone
149 l0| l1| l2| l97| l96 | | l99
155 Here, a set of <b>host</b>s is defined. Each of them has a <b>link</b>
156 to a central backbone (backbone is a link itself, as a link can
157 be used to represent a switch, see the switch / link section
158 below for more details about it). A <b>router</b> allows to connect a
159 <b>cluster</b> to the outside world. Internally,
160 SimGrid treats a cluster as a network zone containing all hosts: the router is the default
161 gateway for the cluster.
163 There is an alternative organization, which is as follows:
177 The principle is the same, except that there is no backbone. This representation
178 can be obtained easily: just do not set the bb_* attributes.
181 Attribute name | Mandatory | Values | Description
182 --------------- | --------- | ------ | -----------
183 id | yes | string | The identifier of the cluster. Facilitates referring to this cluster.
184 prefix | yes | string | Each node of the cluster has to have a name. This name will be prefixed with this prefix.
185 suffix | yes | string | Each node of the cluster will be suffixed with this suffix
186 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.
187 speed | yes | int | Same as the ``speed`` attribute of the ``@<host@>`` tag.
188 core | no | int (default: 1) | Same as the ``core`` attribute of the ``@<host@>`` tag.
189 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the @ref pf_tag_link "link section" for syntax/details.
190 lat | yes | int | Latency for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
191 sharing_policy | no | string | Sharing policy for the links between nodes and backbone (if any). See <b>link</b> section for syntax/details.
192 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>).
193 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>).
194 bb_sharing_policy | no | string | Sharing policy for the backbone (if any). See <b>link</b> section for syntax/details.
195 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 splitduplex 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>
196 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".
197 loopback_lat | no | int | Latency for loopback (if any). See <b>link</b> section for syntax/details. See loopback_bw for more info.
198 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.
199 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".
202 the router name is defined as the resulting String in the following
206 router_name = prefix + clusterId + "_router" + suffix;
210 #### Cluster example ####
212 Consider the following two (and independent) uses of the ``cluster`` tag:
215 <cluster id="my_cluster_1" prefix="" suffix="" radical="0-262144"
216 speed="1e9" bw="125e6" lat="5E-5"/>
218 <cluster id="my_cluster_2" prefix="c-" suffix=".me" radical="0-99"
219 speed="1e9" bw="125e6" lat="5E-5"
220 bb_bw="2.25e9" bb_lat="5E-4"/>
223 The second example creates one router and 100 machines with the following names:
225 c-my_cluster_2_router.me
233 @subsubsection pf_cabinet <cabinet>
236 This tag is only available when the routing mode of the network zone
237 is set to ``Cluster``.
239 The ``<cabinet />`` tag is, like the @ref pf_tag_cluster "<cluster>" tag,
240 a meta-tag. This means that it is simply a shortcut for creating a set of (homogenous) hosts and links quickly;
241 unsurprisingly, this tag was introduced to setup cabinets in data centers quickly. Unlike
242 <cluster>, however, the <cabinet> assumes that you create the backbone
243 and routers yourself; see our examples below.
247 Attribute name | Mandatory | Values | Description
248 --------------- | --------- | ------ | -----------
249 id | yes | string | The identifier of the cabinet. Facilitates referring to this cluster.
250 prefix | yes | string | Each node of the cabinet has to have a name. This name will be prefixed with this prefix.
251 suffix | yes | string | Each node of the cabinet will be suffixed with this suffix
252 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.
253 speed | yes | int | Same as the ``speed`` attribute of the @ref pf_tag_host "<host>" tag.
254 bw | yes | int | Bandwidth for the links between nodes and backbone (if any). See the @ref pf_tag_link "link section" for syntax/details.
255 lat | yes | int | Latency for the links between nodes and backbone (if any). See the @ref pf_tag_link "link section" for syntax/details.
258 Please note that as of now, it is impossible to change attributes such as,
259 amount of cores (always set to 1), the sharing policy of the links (always set to @ref pf_sharing_policy_splitduplex "SPLITDUPLEX").
263 The following example was taken from ``examples/platforms/meta_cluster.xml`` and
264 shows how to use the cabinet tag.
267 <zone id="my_cluster1" routing="Cluster">
268 <cabinet id="cabinet1" prefix="host-" suffix=".cluster1"
269 speed="1Gf" bw="125MBps" lat="100us" radical="1-10"/>
270 <cabinet id="cabinet2" prefix="host-" suffix=".cluster1"
271 speed="1Gf" bw="125MBps" lat="100us" radical="11-20"/>
272 <cabinet id="cabinet3" prefix="host-" suffix=".cluster1"
273 speed="1Gf" bw="125MBps" lat="100us" radical="21-30"/>
275 <backbone id="backbone1" bandwidth="2.25GBps" latency="500us"/>
280 Please note that you must specify the @ref pf_backbone "<backbone>"
281 tag by yourself; this is not done automatically and there are no checks
282 that ensure this backbone was defined.
284 The hosts generated in the above example are named host-1.cluster, host-2.cluster1
288 @subsubsection pf_peer <peer> (Vivaldi netzones only)
290 This tag represents a peer, as in Peer-to-Peer (P2P) networks. This
291 can only be used in Vivaldi NetZones. It creates the following
292 resources to the NetZone:
295 @li Two links: One for download and one for upload. This is
296 convenient to use and simulate stuff under the last mile model (e.g., ADSL peers).
297 @li It connects the two links to the host
301 Attribute name | Mandatory | Values | Description
302 --------------- | --------- | ------ | -----------
303 id | yes | string | The identifier of the peer. Facilitates referring to this peer.
304 speed | yes | int | See the description of the ``host`` tag for this attribute
305 bw_in | yes | int | Bandwidth of the private downstream link
306 bw_out | yes | int | Bandwidth of the private upstream link
307 coordinates | no | string | Coordinates of the gateway for this peer. Example value: 12.8 14.4 6.4
308 sharing_policy | no | SHARED@|SPLITDUPLEX (default: SPLITDUPLEX) | Sharing policy for links. See <b>link</b> description for details.
309 availability_file| no | string | Availability file for the peer. Same as host availability file. See <b>host</b> description for details.
310 state_file | no | string | State file for the peer. Same as host state file. See <b>host</b> description for details.
313 The communication latency between an host A=(xA,yA,zA) and an host
314 B=(xB,yB,zB) is computed as follows:
316 latency = sqrt( (xA-xB)² + (yA-yB)² ) + zA + zB
318 See the documentation of simgrid::kernel::routing::VivaldiZone for
319 details on how the latency is computed from the coordinate, and on the
320 the up and down bandwidth are used.
322 @subsection pf_ne Network equipments
324 There are two tags at all times available to represent network entities and
325 several other tags that are available only in certain contexts.
326 1. ``<link>``: Represents a entity that has a limited bandwidth, a
327 latency, and that can be shared according to TCP way to share this
330 The concept of links in SimGrid may not be intuitive, as links are not
331 limited to connecting (exactly) two entities; in fact, you can have more than
332 two equipments connected to it. (In graph theoretical terms: A link in
333 SimGrid is not an edge, but a hyperedge)
335 2. ``<router/>``: Represents an entity that a message can be routed
336 to, but that is unable to execute any code. In SimGrid, routers have also
337 no impact on the performance: Routers do not limit any bandwidth nor
338 do they increase latency. As a matter of fact, routers are (almost) ignored
339 by the simulator when the simulation has begun.
341 3. ``<backbone/>``: This tag is only available when the containing network zone is
342 used as a cluster (i.e., mode="Cluster")
345 If you want to represent an entity like a switch, you must use ``<link>`` (see section). Routers are used
346 to run some routing algorithm and determine routes (see Section @ref pf_routing for details).
348 @subsubsection pf_router <router/>
350 As said before, <b>router</b> is used only to give some information
351 for routing algorithms. So, it does not have any attributes except:
355 Attribute name | Mandatory | Values | Description
356 --------------- | --------- | ------ | -----------
357 id | yes | string | The identifier of the router to be used when referring to it.
358 coordinates | no | string | Must be provided when choosing the Vivaldi, coordinate-based routing model for the network zone the router belongs to. More details can be found in the Section @ref pf_P2P_tags.
363 <router id="gw_dc1_horizdist"/>
366 @subsubsection pf_tag_link <link>
368 Network links can represent one-hop network connections. They are
369 characterized by their id and their bandwidth; links can (but may not) be subject
374 Attribute name | Mandatory | Values | Description
375 --------------- | --------- | ------ | -----------
376 id | yes | string | The identifier of the link to be used when referring to it.
377 bandwidth | yes | string | Maximum bandwidth for this link, along with its unit.
378 latency | no | double (default: 0.0) | Latency for this link.
379 sharing_policy | no | @ref sharing_policy_shared "SHARED"@|@ref pf_sharing_policy_fatpipe "FATPIPE"@|@ref pf_sharing_policy_splitduplex "SPLITDUPLEX" (default: SHARED) | Sharing policy for the link.
380 bandwidth_file | no | string | Allows you to use a file as input for bandwidth.
381 latency_file | no | string | Allows you to use a file as input for latency.
382 state_file | no | string | Allows you to use a file as input for states.
385 #### Possible shortcuts for ``latency`` ####
387 When using the latency attribute, you can specify the latency by using the scientific
388 notation or by using common abbreviations. For instance, the following three tags
392 <link id="LINK1" bandwidth="125000000" latency="5E-6"/>
393 <link id="LINK1" bandwidth="125000000" latency="5us"/>
394 <link id="LINK1" bandwidth="125000000" latency="0.000005"/>
397 Here, the second tag uses "us", meaning "microseconds". Other shortcuts are:
399 Name | Abbreviation | Time (in seconds)
400 ---- | ------------ | -----------------
401 Week | w | 7 * 24 * 60 * 60
402 Day | d | 24 * 60 * 60
406 Millisecond | ms | 0.001 = 10^(-3)
407 Microsecond | us | 0.000001 = 10^(-6)
408 Nanosecond | ns | 0.000000001 = 10^(-9)
409 Picosecond | ps | 0.000000000001 = 10^(-12)
411 #### Sharing policy ####
413 @anchor sharing_policy_shared
414 By default a network link is @b SHARED, i.e., if two or more data flows go
415 through a link, the bandwidth is shared fairly among all data flows. This
416 is similar to the sharing policy TCP uses.
418 @anchor pf_sharing_policy_fatpipe
419 On the other hand, if a link is defined as a @b FATPIPE,
420 each flow going through this link will be provided with the complete bandwidth,
421 i.e., no sharing occurs and the bandwidth is only limiting each flow individually.
422 Please note that this is really on a per-flow basis, not only on a per-host basis!
423 The complete bandwidth provided by this link in this mode
424 is ``number_of_flows*bandwidth``, with at most ``bandwidth`` being available per flow.
426 Using the FATPIPE mode allows to model backbones that won't affect performance
429 @anchor pf_sharing_policy_splitduplex
430 The last mode available is @b SPLITDUPLEX. This means that SimGrid will
431 automatically generate two links (one carrying the suffix _UP and the other the
432 suffix _DOWN) for each ``<link>`` tag. This models situations when the direction
433 of traffic is important.
436 Transfers from one side to the other will interact similarly as
437 TCP when ACK returning packets circulate on the other direction. More
438 discussion about it is available in the description of link_ctn description.
440 In other words: The SHARED policy defines a physical limit for the bandwidth.
441 The FATPIPE mode defines a limit for each application,
442 with no upper total limit.
445 Tip: By using the FATPIPE mode, you can model big backbones that
446 won't affect performance (except latency).
451 <link id="SWITCH" bandwidth="125000000" latency="5E-5" sharing_policy="FATPIPE" />
454 #### Expressing dynamism and failures ####
456 Similar to hosts, it is possible to declare links whose state, bandwidth
457 or latency changes over time (see Section @ref pf_host_dynamism for details).
459 In the case of network links, the ``bandwidth`` and ``latency`` attributes are
460 replaced by the ``bandwidth_file`` and ``latency_file`` attributes.
461 The following XML snippet demonstrates how to use this feature in the platform
462 file. The structure of the files "link1.bw" and "link1.lat" is shown below.
465 <link id="LINK1" state_file="link1.fail" bandwidth="80000000" latency=".0001" bandwidth_file="link1.bw" latency_file="link1.lat" />
469 Even if the syntax is the same, the semantic of bandwidth and latency
470 trace files differs from that of host availability files. For bandwidth and
471 latency, the corresponding files do not
472 express availability as a fraction of the available capacity but directly in
473 bytes per seconds for the bandwidth and in seconds for the latency. This is
474 because most tools allowing to capture traces on real platforms (such as NWS)
475 express their results this way.
477 ##### Example of "link1.bw" file #####
485 In this example, the bandwidth changes repeatedly, with all changes
486 being repeated every 12 seconds.
488 At the beginning of the the simulation, the link's bandwidth is 80,000,000
489 B/s (i.e., 80 Mb/s); this value was defined in the XML snippet above.
490 After four seconds, it drops to 40 Mb/s (line 2), and climbs
491 back to 60 Mb/s after another 4 seconds (line 3). The value does not change any
492 more until the end of the period, that is, after 12 seconds have been simulated).
493 At this point, periodicity kicks in and this behavior is repeated: Seconds
494 12-16 will experience 80 Mb/s, 16-20 40 Mb/s etc.).
496 ##### Example of "link1.lat" file #####
505 In this example, the latency varies with a period of 5 seconds.
506 In the xml snippet above, the latency is initialized to be 0.0001s (100µs). This
507 value will be kept during the first second, since the latency_file contains
508 changes to this value at second one, two and three.
509 At second one, the value will be 0.001, i.e., 1ms. One second later it will
510 be adjusted to 0.01 (or 10ms) and one second later it will be set again to 1ms. The
511 value will not change until second 5, when the periodicity defined in line 1
512 kicks in. It then loops back, starting at 100µs (the initial value) for one second.
514 #### The ``<prop/>`` tag ####
516 Similar to the ``<host>`` tag, a link may also contain the ``<prop/>`` tag; see the host
517 documentation (Section @ref pf_tag_host) for an example.
520 @subsubsection pf_backbone <backbone/>
523 This tag is <b>only available</b> when the containing network zone uses the "Cluster" routing mode!
525 Using this tag, you can designate an already existing link to be a backbone.
527 Attribute name | Mandatory | Values | Description
528 --------------- | --------- | ------ | -----------
529 id | yes | string | Name of the link that is supposed to act as a backbone.
531 @subsection pf_storage Storage
534 This is a prototype version that should evolve quickly, hence this
535 is just some doc valuable only at the time of writing.
536 This section describes the storage management under SimGrid ; nowadays
537 it's only usable with MSG. It relies basically on linux-like concepts.
538 You also may want to have a look to its corresponding section in
539 @ref msg_file ; access functions are organized as a POSIX-like
542 @subsubsection pf_sto_conc Storage - Main Concepts
544 The storage facilities implemented in SimGrid help to model (and account for)
545 storage devices, such as tapes, hard-drives, CD or DVD devices etc.
546 A typical situation is depicted in the figure below:
548 @image html ./webcruft/storage_sample_scenario.png
549 @image latex ./webcruft/storage_sample_scenario.png "storage_sample_scenario" width=@textwidth
551 In this figure, two hosts called Bob and Alice are interconnected via a network
552 and each host is physically attached to a disk; it is not only possible for each host to
553 mount the disk they are attached to directly, but they can also mount disks
554 that are in a remote location. In this example, Bob mounts Alice's disk remotely
555 and accesses the storage via the network.
557 SimGrid provides 3 different entities that can be used to model setups
558 that include storage facilities:
560 Entity name | Description
561 --------------- | -----------
562 @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).
563 @ref pf_tag_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.
564 @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).
567 @anchor pf_storage_content_file
568 ### %Storage Content File ###
570 In order to assess exactly how much time is spent reading from the storage,
571 SimGrid needs to know what is stored on the storage device (identified by distinct (file-)name, like in a file system)
572 and what size this content has.
575 The content file is never changed by the simulation; it is parsed once
576 per simulation and kept in memory afterwards. When the content of the
577 storage changes, only the internal SimGrid data structures change.
579 @anchor pf_storage_content_file_structure
580 #### Structure of a %Storage Content File ####
582 Here is an excerpt from two storage content file; if you want to see the whole file, check
583 the file ``examples/platforms/content/storage_content.txt`` that comes with the
586 SimGrid essentially supports two different formats: UNIX-style filepaths should
587 follow the well known format:
590 /lib/libsimgrid.so.3.6.2 12710497
594 /bin/simgrid_update_xml 5018
595 /bin/graphicator 66986
596 /bin/simgrid-colorizer 2993
601 Windows filepaths, unsurprisingly, use the windows style:
604 @Windows@avastSS.scr 41664
605 @Windows@bfsvc.exe 75264
606 @Windows@bootstat.dat 67584
607 @Windows@CoreSingleLanguage.xml 31497
609 @Windows@dchcfg64.exe 335464
610 @Windows@dcmdev64.exe 93288
614 The different file formats come at a cost; in version 3.12 (and most likely
615 in later versions, too), copying files from windows-style storages to unix-style
616 storages (and vice versa) is not supported.
618 @anchor pf_storage_content_file_create
619 #### Generate a %Storage Content File ####
621 If you want to generate a storage content file based on your own filesystem (or at least a filesystem you have access to),
622 try running this command (works only on unix systems):
625 find . -type f -exec ls -1s --block=1 {} @; 2>/dev/null | awk '{ print $2 " " $1}' > ./content.txt
628 @subsubsection pf_storage_entities The Storage Entities
630 These are the entities that you can use in your platform files to include
631 storage in your model. See also the list of our @ref pf_storage_example_files "example files";
632 these might also help you to get started.
634 @anchor pf_storage_entity_storage_type
635 #### @<storage_type@> ####
637 Attribute name | Mandatory | Values | Description
638 --------------- | --------- | ------ | -----------
639 id | yes | string | Identifier of this storage_type; used when referring to it
640 model | no | string | In the future, this will allow to change the performance model to use
641 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)
642 content | yes | string | Path to a @ref pf_storage_content_file "Storage Content File" on your system. This file must exist.
644 This tag must contain some predefined model properties, specified via the <model_prop> tag. Here is a list,
645 see below for an example:
647 Property id | Mandatory | Values | Description
648 --------------- | --------- | ------ | -----------
649 Bwrite | yes | string | Bandwidth for write access; in B/s (but you can also specify e.g. "30MBps")
650 Bread | yes | string | Bandwidth for read access; in B/s (but you can also specify e.g. "30MBps")
653 A storage_type can also contain the <b><prop></b> tag. The <prop> tag allows you
654 to associate additional information to this <storage_type> and follows the
655 attribute/value schema; see the example below. You may want to use it to give information to
656 the tool you use for rendering your simulation, for example.
658 Here is a complete example for the ``storage_type`` tag:
660 <storage_type id="single_HDD" size="4000">
661 <model_prop id="Bwrite" value="30MBps" />
662 <model_prop id="Bread" value="100MBps" />
663 <prop id="Brand" value="Western Digital" />
667 @subsubsection pf_tag_storage <storage>
669 Attributes | Mandatory | Values | Description
670 -------------- | --------- | ------ | -----------
671 id | yes | string | Identifier of this ``storage``; used when referring to it
672 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.
673 attach | yes | string | Name of a host (see Section @ref pf_tag_host) to which this storage is <i>physically</i> attached to (e.g., a hard drive in a computer)
674 content | no | string | When specified, overwrites the content attribute of @ref pf_storage_entity_storage_type "@<storage_type@>"
676 Here are two examples:
679 <storage id="Disk1" typeId="single_HDD" attach="bob" />
681 <storage id="Disk2" typeId="single_SSD"
682 content="content/win_storage_content.txt" />
685 The first example is straightforward: A disk is defined and called "Disk1"; it is
686 of type "single_HDD" (shown as an example of @ref pf_storage_entity_storage_type "@<storage_type@>" above) and attached
687 to a host called "bob" (the definition of this host is omitted here).
689 The second storage is called "Disk2", is still of the same type as Disk1 but
690 now specifies a new content file (so the contents will be different from Disk1)
691 and the filesystem uses the windows style; finally, it is attached to a second host,
692 called alice (which is again not defined here).
694 @subsubsection pf_tag_mount <mount>
696 | Attribute | Mandatory | Values | Description |
697 | ----------- | ----------- | -------- | ------------- |
698 | id | yes | string | Refers to a @ref pf_tag_storage "<storage>" entity that will be mounted on that computer |
699 | name | yes | string | Path/location to/of the logical reference (mount point) of this disk
701 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)
702 is; this location is specified by the ``name`` attribute.
704 Here is a simple example, taken from the file ``examples/platform/storage.xml``:
707 <storage_type id="single_SSD" size="500GiB">
708 <model_prop id="Bwrite" value="60MBps" />
709 <model_prop id="Bread" value="200MBps" />
712 <storage id="Disk2" typeId="single_SSD"
713 content="content/win_storage_content.txt"
715 <storage id="Disk4" typeId="single_SSD"
716 content="content/small_content.txt"
719 <host id="alice" speed="1Gf">
720 <mount storageId="Disk2" name="c:"/>
723 <host id="denise" speed="1Gf">
724 <mount storageId="Disk2" name="c:"/>
725 <mount storageId="Disk4" name="/home"/>
729 This example is quite interesting, as the same device, called "Disk2", is mounted by
730 two hosts at the same time! Note, however, that the host called ``alice`` is actually
731 attached to this storage, as can be seen in the @ref pf_tag_storage "<storage>"
732 tag. This means that ``denise`` must access this storage through the network, but SimGrid automatically takes
733 care of that for you.
735 Furthermore, this example shows that ``denise`` has mounted two storages with different
736 filesystem types (unix and windows). In general, a host can mount as many storage devices as
740 Again, the difference between ``attach`` and ``mount`` is simply that
741 an attached storage is always physically inside (or connected to) that machine;
742 for instance, a USB stick is attached to one and only one machine (where it's plugged-in)
743 but it can only be mounted on others, as mounted storage can also be a remote location.
745 ###### Example files #####
747 @verbinclude example_filelist_xmltag_mount
749 @subsubsection pf_storage_example_files Example files
751 Several examples were already discussed above; if you're interested in full examples,
752 check the the following platforms:
754 1. ``examples/platforms/storage.xml``
755 2. ``examples/platforms/remote_io.xml``
757 If you're looking for some examplary C code, you may find the source code
758 available in the directory ``examples/msg/io/`` useful.
760 @subsubsection pf_storage_examples_modelling Modelling different situations
762 The storage functionality of SimGrid is type-agnostic, that is, the implementation
763 does not presume any type of storage, such as HDDs/SSDs, RAM,
764 CD/DVD devices, USB sticks etc.
766 This allows the user to apply the simulator for a wide variety of scenarios; one
767 common scenario would be the access of remote RAM.
769 #### Modelling the access of remote RAM ####
771 How can this be achieved in SimGrid? Let's assume we have a setup where three hosts
772 (HostA, HostB, HostC) need to access remote RAM:
782 An easy way to model this scenario is to setup and define the RAM via the
783 @ref pf_tag_storage "storage" and @ref pf_storage_entity_storage_type "storage type"
784 entities and attach it to a remote dummy host; then, every host can have their own links
785 to this host (modelling for instance certain scenarios, such as PCIe ...)
790 RAM - Dummy -- Host B
795 Now, if read from this storage, the host that mounts this storage
796 communicates to the dummy host which reads from RAM and
797 sends the information back.
800 @section pf_routing Routing
802 To achieve high performance, the routing tables used within SimGrid are
803 static. This means that routing between two nodes is calculated once
804 and will not change during execution. The SimGrid team chose to use this
805 approach as it is rare to have a real deficiency of a resource;
806 most of the time, a communication fails because the links experience too much
807 congestion and hence, your connection stops before the timeout or
808 because the computer designated to be the destination of that message
811 We also chose to use shortest paths algorithms in order to emulate
812 routing. Doing so is consistent with the reality: [RIP](https://en.wikipedia.org/wiki/Routing_Information_Protocol),
813 [OSPF](https://en.wikipedia.org/wiki/Open_Shortest_Path_First), [BGP](https://en.wikipedia.org/wiki/Border_Gateway_Protocol)
814 are all calculating shortest paths. They do require some time to converge, but
815 eventually, when the routing tables have stabilized, your packets will follow
818 @subsection pf_tag_zone <zone>
820 Before SimGrid v3.16, networking zones used to be called Autonomous
821 Systems, but this was misleading as zones may include other zones in a
822 hierarchical manner. If you find any remaining reference to network
823 zones, please report this as a bug.
825 Attribute | Value | Description
826 ----------- | ------------------------------------------------- | ----------------------------------------------
827 id | String (mandatory) | The identifier of this zone (must be unique)
828 routing | One of the existing routing algorithm (mandatory) | See Section @ref pf_rm for details.
832 <zone id="zone0" routing="Full">
833 <host id="host1" speed="1000000000"/>
834 <host id="host2" speed="1000000000"/>
835 <link id="link1" bandwidth="125000000" latency="0.000100"/>
836 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
840 In this example, zone0 contains two hosts (host1 and host2). The route
841 between the hosts goes through link1.
843 @subsection pf_rm Routing models
845 For each network zone, you must define explicitly which routing model will
846 be used. There are 3 different categories for routing models:
848 1. @ref pf_routing_model_shortest_path "Shortest-path" based models: SimGrid calculates shortest
849 paths and manages them. Behaves more or less like most real life
851 2. @ref pf_routing_model_manual "Manually-entered" route models: you have to define all routes
852 manually in the platform description file; this can become
853 tedious very quickly, as it is very verbose.
854 Consistent with some manually managed real life routing.
855 3. @ref pf_routing_model_simple "Simple/fast models": those models offer fast, low memory routing
856 algorithms. You should consider to use this type of model if
857 you can make some assumptions about your network zone.
858 Routing in this case is more or less ignored.
860 @subsubsection pf_raf The router affair
862 Using routers becomes mandatory when using shortest-path based
863 models or when using the bindings to the ns-3 packet-level
864 simulator instead of the native analytical network model implemented
867 For graph-based shortest path algorithms, routers are mandatory, because these
868 algorithms require a graph as input and so we need to have source and
869 destination for each edge.
871 Routers are naturally an important concept ns-3 since the
872 way routers run the packet routing algorithms is actually simulated.
873 SimGrid's analytical models however simply aggregate the routing time
874 with the transfer time.
876 So why did we incorporate routers in SimGrid? Rebuilding a graph representation
877 only from the route information turns out to be a very difficult task, because
878 of the missing information about how routes intersect. That is why we
879 introduced routers, which are simply used to express these intersection points.
880 It is important to understand that routers are only used to provide topological
883 To express this topological information, a <b>route</b> has to be
884 defined in order to declare which link is connected to a router.
887 @subsubsection pf_routing_model_shortest_path Shortest-path based models
889 The following table shows all the models that compute routes using
890 shortest-paths algorithms are currently available in SimGrid. More detail on how
891 to choose the best routing model is given in the Section called @"@ref pf_routing_howto_choose_wisely@".
893 | Name | Description |
894 | --------------------------------------------------- | -------------------------------------------------------------------------- |
895 | @ref pf_routing_model_floyd "Floyd" | Floyd routing data. Pre-calculates all routes once |
896 | @ref pf_routing_model_dijkstra "Dijkstra" | Dijkstra routing data. Calculates routes only when needed |
897 | @ref pf_routing_model_dijkstracache "DijkstraCache" | Dijkstra routing data. Handles some cache for already calculated routes. |
899 All those shortest-path models are instanciated in the same way and are
900 completely interchangeable. Here are some examples:
902 @anchor pf_routing_model_floyd
907 <zone id="zone0" routing="Floyd">
909 <cluster id="my_cluster_1" prefix="c-" suffix=""
910 radical="0-1" speed="1000000000" bw="125000000" lat="5E-5"
911 router_id="router1"/>
913 <zone id="zone1" routing="None">
914 <host id="host1" speed="1000000000"/>
917 <link id="link1" bandwidth="100000" latency="0.01"/>
919 <zoneroute src="my_cluster_1" dst="zone1"
922 <link_ctn id="link1"/>
928 zoneroute given at the end gives a topological information: link1 is
929 between router1 and host1.
931 #### Example platform files ####
933 This is an automatically generated list of example files that use the Floyd
934 routing model (the path is given relative to SimGrid's source directory)
936 @verbinclude example_filelist_routing_floyd
938 @anchor pf_routing_model_dijkstra
941 #### Example platform files ####
943 This is an automatically generated list of example files that use the Dijkstra
944 routing model (the path is given relative to SimGrid's source directory)
946 @verbinclude example_filelist_routing_dijkstra
950 <zone id="zone_2" routing="Dijkstra">
951 <host id="zone_2_host1" speed="1000000000"/>
952 <host id="zone_2_host2" speed="1000000000"/>
953 <host id="zone_2_host3" speed="1000000000"/>
954 <link id="zone_2_link1" bandwidth="1250000000" latency="5E-4"/>
955 <link id="zone_2_link2" bandwidth="1250000000" latency="5E-4"/>
956 <link id="zone_2_link3" bandwidth="1250000000" latency="5E-4"/>
957 <link id="zone_2_link4" bandwidth="1250000000" latency="5E-4"/>
958 <router id="central_router"/>
959 <router id="zone_2_gateway"/>
960 <!-- routes providing topological information -->
961 <route src="central_router" dst="zone_2_host1"><link_ctn id="zone_2_link1"/></route>
962 <route src="central_router" dst="zone_2_host2"><link_ctn id="zone_2_link2"/></route>
963 <route src="central_router" dst="zone_2_host3"><link_ctn id="zone_2_link3"/></route>
964 <route src="central_router" dst="zone_2_gateway"><link_ctn id="zone_2_link4"/></route>
968 @anchor pf_routing_model_dijkstracache
969 ### DijkstraCache ###
971 DijkstraCache example:
973 <zone id="zone_2" routing="DijkstraCache">
974 <host id="zone_2_host1" speed="1000000000"/>
976 (platform unchanged compared to upper example)
979 #### Example platform files ####
981 This is an automatically generated list of example files that use the DijkstraCache
982 routing model (the path is given relative to SimGrid's source directory):
984 Editor's note: At the time of writing, no platform file used this routing model - so
985 if there are no example files listed here, this is likely to be correct.
987 @verbinclude example_filelist_routing_dijkstra_cache
989 @subsubsection pf_routing_model_manual Manually-entered route models
991 | Name | Description |
992 | ---------------------------------- | ------------------------------------------------------------------------------ |
993 | @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. |
995 @anchor pf_routing_model_full
1000 <zone id="zone0" routing="Full">
1001 <host id="host1" speed="1000000000"/>
1002 <host id="host2" speed="1000000000"/>
1003 <link id="link1" bandwidth="125000000" latency="0.000100"/>
1004 <route src="host1" dst="host2"><link_ctn id="link1"/></route>
1008 #### Example platform files ####
1010 This is an automatically generated list of example files that use the Full
1011 routing model (the path is given relative to SimGrid's source directory):
1013 @verbinclude example_filelist_routing_full
1015 @subsubsection pf_routing_model_simple Simple/fast models
1017 | Name | Description |
1018 | ---------------------------------------- | ------------------------------------------------------------------------------ |
1019 | @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. |
1020 | @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. |
1021 | @ref pf_routing_model_vivaldi "Vivaldi" | Perfect when you want to use coordinates. Also see the corresponding @ref pf_P2P_tags "P2P section" below. |
1023 @anchor pf_routing_model_cluster
1027 In this mode, the @ref pf_cabinet "<cabinet/>" tag is available.
1029 #### Example platform files ####
1031 This is an automatically generated list of example files that use the Cluster
1032 routing model (the path is given relative to SimGrid's source directory):
1034 @verbinclude example_filelist_routing_cluster
1036 @anchor pf_routing_model_none
1040 This model does exactly what it's name advertises: Nothing. There is no routing
1041 available within this model and if you try to communicate within the zone that
1042 uses this model, SimGrid will fail unless you have explicitly activated the
1043 @ref options_model_select_network_constant "Constant Network Model" (this model charges
1044 the same for every single communication). It should
1045 be noted, however, that you can still attach an @ref pf_tag_zoneroute "ZoneRoute",
1046 as is demonstrated in the example below:
1048 @verbinclude platforms/cluster_and_one_host.xml
1050 #### Example platform files ####
1052 This is an automatically generated list of example files that use the None
1053 routing model (the path is given relative to SimGrid's source directory):
1055 @verbinclude example_filelist_routing_none
1058 @anchor pf_routing_model_vivaldi
1061 For more information on how to use the [Vivaldi Coordinates](https://en.wikipedia.org/wiki/Vivaldi_coordinates),
1062 see also Section @ref pf_P2P_tags "P2P tags".
1064 Note that it is possible to combine the Vivaldi routing model with other routing models;
1065 an example can be found in the file @c examples/platforms/cloud.xml. This
1066 examples models a NetZone using Vivaldi that contains other NetZones that use different
1069 #### Example platform files ####
1071 This is an automatically generated list of example files that use the None
1072 routing model (the path is given relative to SimGrid's source directory):
1074 @verbinclude example_filelist_routing_vivaldi
1077 @subsection ps_dec Defining routes
1079 There are currently four different ways to define routes:
1081 | Name | Description |
1082 | ------------------------------------------------- | ----------------------------------------------------------------------------------- |
1083 | @ref pf_tag_route "route" | Used to define route between host/router |
1084 | @ref pf_tag_zoneroute "zoneRoute" | Used to define route between different zones |
1085 | @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. |
1086 | @ref pf_tag_bypassasroute "bypassZoneRoute" | Used in the same way as bypassRoute, but for zones |
1088 Basically all those tags will contain an (ordered) list of references
1089 to link that compose the route you want to define.
1091 Consider the example below:
1094 <route src="Alice" dst="Bob">
1095 <link_ctn id="link1"/>
1096 <link_ctn id="link2"/>
1097 <link_ctn id="link3"/>
1101 The route here from host Alice to Bob will be first link1, then link2,
1102 and finally link3. What about the reverse route? @ref pf_tag_route "Route" and
1103 @ref pf_tag_zoneroute "zoneroute" have an optional attribute @c symmetrical, that can
1104 be either @c YES or @c NO. @c YES means that the reverse route is the same
1105 route in the inverse order, and is set to @c YES by default. Note that
1106 this is not the case for bypass*Route, as it is more probable that you
1107 want to bypass only one default route.
1109 For an @ref pf_tag_zoneroute "zoneroute", things are just slightly more complicated, as you have
1110 to give the id of the gateway which is inside the zone you want to access ...
1111 So it looks like this:
1114 <zoneroute src="zone1" dst="zone2"
1115 gw_src="router1" gw_dst="router2">
1116 <link_ctn id="link1"/>
1120 gw == gateway, so when any message are trying to go from zone1 to zone2,
1121 it means that it must pass through router1 to get out of the zone, then
1122 pass through link1, and get into zone2 by being received by router2.
1123 router1 must belong to zone1 and router2 must belong to zone2.
1125 @subsubsection pf_tag_linkctn <link_ctn>
1127 This entity has only one purpose: Refer to an already existing
1128 @ref pf_tag_link "<link/>" when defining a route, i.e., it
1129 can only occur as a child of @ref pf_tag_route "<route/>"
1131 | Attribute name | Mandatory | Values | Description |
1132 | --------------- | --------- | ------ | ----------- |
1133 | id | yes | String | The identifier of the link that should be added to the route. |
1134 | direction | maybe | UP@|DOWN | If the link referenced by @c id has been declared as @ref pf_sharing_policy_splitduplex "SPLITDUPLEX", this indicates which direction the route traverses through this link: UP or DOWN. If you don't use SPLITDUPLEX, do not use this attribute or SimGrid will not find the right link.
1136 #### Example Files ####
1138 This is an automatically generated list of example files that use the @c <link_ctn/>
1139 entity (the path is given relative to SimGrid's source directory):
1141 @verbinclude example_filelist_xmltag_linkctn
1143 @subsubsection pf_tag_zoneroute <zoneRoute>
1145 The purpose of this entity is to define a route between two
1146 NetZones. Recall that all zones form a tree, so to connect two
1147 sibiling zones, you must give such a zoneRoute specifying the source
1148 and destination zones, along with the gateway in each zone (ie, the
1149 point to reach within that zone to reach the netzone), and the list of
1150 links in the ancestor zone to go from one zone to another.
1152 So, to go from an host @c src_host that is within zone @c src, to an
1153 host @c dst_host that is within @c dst, you need to:
1155 - move within zone @c src, from @c src_host to the specified @c gw_src;
1156 - traverse all links specified by the zoneRoute (they are supposed to be within the common ancestor);
1157 - move within zone @c dst, from @c gw_dst to @c dst_host.
1159 #### Attributes ####
1161 | Attribute name | Mandatory | Values | Description |
1162 | --------------- | --------- | ------ | ----------- |
1163 | src | yes | String | The identifier of the source zone |
1164 | dst | yes | String | See the @c src attribute |
1165 | gw_src | yes | String | The gateway that will be used within the src zone; this can be any @ref pf_tag_host "Host" or @ref pf_router "Router" defined within the src zone. |
1166 | gw_dst | yes | String | Same as @c gw_src, but with the dst zone instead. |
1167 | symmetrical | no | YES@|NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1172 <zone id="zone0" routing="Full">
1173 <cluster id="my_cluster_1" prefix="c-" suffix=".me"
1174 radical="0-149" speed="1000000000" bw="125000000" lat="5E-5"
1175 bb_bw="2250000000" bb_lat="5E-4"/>
1177 <cluster id="my_cluster_2" prefix="c-" suffix=".me"
1178 radical="150-299" speed="1000000000" bw="125000000" lat="5E-5"
1179 bb_bw="2250000000" bb_lat="5E-4"/>
1181 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1183 <zoneroute src="my_cluster_1" dst="my_cluster_2"
1184 gw_src="c-my_cluster_1_router.me"
1185 gw_dst="c-my_cluster_2_router.me">
1186 <link_ctn id="backbone"/>
1188 <zoneroute src="my_cluster_2" dst="my_cluster_1"
1189 gw_src="c-my_cluster_2_router.me"
1190 gw_dst="c-my_cluster_1_router.me">
1191 <link_ctn id="backbone"/>
1196 @subsubsection pf_tag_route <route>
1198 The principle is the same as for
1199 @ref pf_tag_zoneroute "ZoneRoute": The route contains a list of links that
1200 provide a path from @c src to @c dst. Here, @c src and @c dst can both be either a
1201 @ref pf_tag_host "host" or @ref pf_router "router". This is mostly useful for the
1202 @ref pf_routing_model_full "Full routing model" as well as for the
1203 @ref pf_routing_model_shortest_path "shortest-paths" based models (as they require
1204 topological information).
1207 | Attribute name | Mandatory | Values | Description |
1208 | --------------- | --------- | ---------------------- | ----------- |
1209 | src | yes | String | The value given to the source's "id" attribute |
1210 | dst | yes | String | The value given to the destination's "id" attribute. |
1211 | symmetrical | no | YES@| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1216 A route in the @ref pf_routing_model_full "Full routing model" could look like this:
1218 <route src="Tremblay" dst="Bourassa">
1219 <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"/>
1223 A route in the @ref pf_routing_model_shortest_path "Shortest-Path routing model" could look like this:
1225 <route src="Tremblay" dst="Bourassa">
1230 You must only have one link in your routes when you're using them to provide
1231 topological information, as the routes here are simply the edges of the
1232 (network-)graph and the employed algorithms need to know which edge connects
1233 which pair of entities.
1235 @subsubsection pf_tag_bypassasroute bypasszoneroute
1237 As said before, once you choose
1238 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1239 define some of your routes, which will be specific. You may also want
1240 to bypass some routes defined in lower level zone at an upper stage:
1241 <b>bypasszoneroute</b> is the tag you're looking for. It allows to
1242 bypass routes defined between already defined between zone (if you want
1243 to bypass route for a specific host, you should just use byPassRoute).
1244 The principle is the same as zoneroute: <b>bypasszoneroute</b> contains
1245 list of links that are in the path between src and dst.
1247 #### Attributes ####
1249 | Attribute name | Mandatory | Values | Description |
1250 | --------------- | --------- | ---------------------- | ----------- |
1251 | src | yes | String | The value given to the source zone's "id" attribute |
1252 | dst | yes | String | The value given to the destination zone's "id" attribute. |
1253 | gw_src | yes | String | The value given to the source gateway's "id" attribute; this can be any host or router within the src zone |
1254 | gw_dst | yes | String | The value given to the destination gateway's "id" attribute; this can be any host or router within the dst zone|
1255 | symmetrical | no | YES@| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1260 <bypasszoneRoute src="my_cluster_1" dst="my_cluster_2"
1261 gw_src="my_cluster_1_router"
1262 gw_dst="my_cluster_2_router">
1263 <link_ctn id="link_tmp"/>
1267 This example shows that link @c link_tmp (definition not displayed here) directly
1268 connects the router @c my_cluster_1_router in the source cluster to the router
1269 @c my_cluster_2_router in the destination router. Additionally, as the @c symmetrical
1270 attribute was not given, this route is presumed to be symmetrical.
1272 @subsubsection pf_tag_bypassroute bypassRoute
1274 As said before, once you choose
1275 a model, it (most likely; the constant network model, for example, doesn't) calculates routes for you. But maybe you want to
1276 define some of your routes, which will be specific. You may also want
1277 to bypass some routes defined in lower level zone at an upper stage:
1278 <b>bypassRoute</b> is the tag you're looking for. It allows to bypass
1279 routes defined between <b>host/router</b>. The principle is the same
1280 as route: <b>bypassRoute</b> contains list of links references of
1281 links that are in the path between src and dst.
1283 #### Attributes ####
1285 | Attribute name | Mandatory | Values | Description |
1286 | --------------- | --------- | ---------------------- | ----------- |
1287 | src | yes | String | The value given to the source zone's "id" attribute |
1288 | dst | yes | String | The value given to the destination zone's "id" attribute. |
1289 | symmetrical | no | YES @| NO (Default: YES) | If this route is symmetric, the opposite route (from dst to src) will also be declared implicitly. |
1294 <bypassRoute src="host_1" dst="host_2">
1295 <link_ctn id="link_tmp"/>
1299 This example shows that link @c link_tmp (definition not displayed here) directly
1300 connects host @c host_1 to host @c host_2. Additionally, as the @c symmetrical
1301 attribute was not given, this route is presumed to be symmetrical.
1303 @subsection pb_baroex Basic Routing Example
1305 Let's say you have an zone named zone_Big that contains two other zone, zone_1
1306 and zone_2. If you want to make a host (h1) from zone_1 with another one
1307 (h2) from zone_2 then you'll have to proceed as follows:
1308 @li First, you have to ensure that a route is defined from h1 to the
1309 zone_1's exit gateway and from h2 to zone_2's exit gateway.
1310 @li Then, you'll have to define a route between zone_1 to zone_2. As those
1311 zone are both resources belonging to zone_Big, then it has to be done
1312 at zone_big level. To define such a route, you have to give the
1313 source zone (zone_1), the destination zone (zone_2), and their respective
1314 gateway (as the route is effectively defined between those two
1315 entry/exit points). Elements of this route can only be elements
1316 belonging to zone_Big, so links and routers in this route should be
1317 defined inside zone_Big. If you choose some shortest-path model,
1318 this route will be computed automatically.
1320 As said before, there are mainly 2 tags for routing:
1321 @li <b>zoneroute</b>: to define routes between two <b>zone</b>
1322 @li <b>route</b>: to define routes between two <b>host/router</b>
1324 As we are dealing with routes between zone, it means that those we'll
1325 have some definition at zone_Big level. Let consider zone_1 contains 1
1326 host, 1 link and one router and zone_2 3 hosts, 4 links and one router.
1327 There will be a central router, and a cross-like topology. At the end
1328 of the crosses arms, you'll find the 3 hosts and the router that will
1329 act as a gateway. We have to define routes inside those two zone. Let
1330 say that zone_1 contains full routes, and zone_2 contains some Floyd
1331 routing (as we don't want to bother with defining all routes). As
1332 we're using some shortest path algorithms to route into zone_2, we'll
1333 then have to define some <b>route</b> to gives some topological
1334 information to SimGrid. Here is a file doing it all:
1337 <zone id="zone_Big" routing="Dijkstra">
1338 <zone id="zone_1" routing="Full">
1339 <host id="zone_1_host1" speed="1000000000"/>
1340 <link id="zone_1_link" bandwidth="1250000000" latency="5E-4"/>
1341 <router id="zone_1_gateway"/>
1342 <route src="zone_1_host1" dst="zone_1_gateway">
1343 <link_ctn id="zone_1_link"/>
1346 <zone id="zone_2" routing="Floyd">
1347 <host id="zone_2_host1" speed="1000000000"/>
1348 <host id="zone_2_host2" speed="1000000000"/>
1349 <host id="zone_2_host3" speed="1000000000"/>
1350 <link id="zone_2_link1" bandwidth="1250000000" latency="5E-4"/>
1351 <link id="zone_2_link2" bandwidth="1250000000" latency="5E-4"/>
1352 <link id="zone_2_link3" bandwidth="1250000000" latency="5E-4"/>
1353 <link id="zone_2_link4" bandwidth="1250000000" latency="5E-4"/>
1354 <router id="central_router"/>
1355 <router id="zone_2_gateway"/>
1356 <!-- routes providing topological information -->
1357 <route src="central_router" dst="zone_2_host1"><link_ctn id="zone_2_link1"/></route>
1358 <route src="central_router" dst="zone_2_host2"><link_ctn id="zone_2_link2"/></route>
1359 <route src="central_router" dst="zone_2_host3"><link_ctn id="zone_2_link3"/></route>
1360 <route src="central_router" dst="zone_2_gateway"><link_ctn id="zone_2_link4"/></route>
1362 <link id="backbone" bandwidth="1250000000" latency="5E-4"/>
1364 <zoneroute src="zone_1" dst="zone_2"
1365 gw_src="zone_1_gateway"
1366 gw_dst="zone_2_gateway">
1367 <link_ctn id="backbone"/>
1372 @section pf_other Other tags
1374 The following tags can be used inside a @<platform@> tag even if they are not
1375 directly describing the platform:
1377 - @ref pf_tag_config passes configuration options, e.g. to change the network model;
1378 - @ref pf_tag_prop gives user-defined properties to various elements
1380 @subsection pf_tag_config <config>
1382 Adding configuration flags into the platform file is particularly
1383 useful when the described platform is best used with specific
1384 flags. For example, you could finely tune SMPI in your platform file directly.
1386 | Attribute | Values | Description |
1387 | ---------- | ------------------- | ---------------------------------------------- |
1388 | id | String (optional) | This optional identifier is ignored by SimGrid |
1390 * **Included tags:** @ref pf_tag_prop to specify a given configuration item (see @ref options).
1392 Any such configuration must be given at the very top of the platform file.
1397 <?xml version='1.0'?>
1398 <!DOCTYPE platform SYSTEM "https://simgrid.org/simgrid.dtd">
1399 <platform version="4">
1401 <prop id="maxmin/precision" value="0.000010" />
1402 <prop id="cpu/optim" value="TI" />
1403 <prop id="network/model" value="SMPI" />
1404 <prop id="smpi/bw-factor" value="65472:0.940694;15424:0.697866;9376:0.58729" />
1407 <zone id="zone0" routing="Full">
1411 @subsection pf_tag_prop <prop>
1413 Defines a user-defined property, identified with a name and having a
1414 value. You can specify such properties to most kind of resources:
1415 @ref pf_tag_zone, @ref pf_tag_host, @ref pf_tag_storage,
1416 @ref pf_tag_cluster and @ref pf_tag_link. These values can be retrieved
1417 at runtime with MSG_zone_property() or simgrid::s4u::NetZone::property(),
1418 or similar functions.
1420 | Attribute | Values | Description |
1421 | --------- | ---------------------- | ----------------------------------------------------------------------------------------- |
1422 | id | String (mandatory) | Identifier of this property. Must be unique for a given property holder, eg host or link. |
1423 | value | String (mandatory) | Value of this property; The semantic is completely up to you. |
1425 * **Included tags:** none.
1430 <prop id="Operating System" value="Linux" />
1434 @subsection pf_trace trace and trace_connect
1436 Both tags are an alternate way to pass files containing information on
1437 availability, state etc. to an entity. (See also @ref howto_churn).
1438 Instead of referring to the file directly in the host, link, or
1439 cluster tag, you proceed by defining a trace with an id corresponding
1440 to a file, later a host/link/cluster, and finally using trace_connect
1441 you say that the file trace must be used by the entity.
1447 <zone id="zone0" routing="Full">
1448 <host id="bob" speed="1000000000"/>
1450 <trace id="myTrace" file="bob.trace" periodicity="1.0"/>
1451 <trace_connect trace="myTrace" element="bob" kind="POWER"/>
1455 The order here is important. @c trace_connect must come
1456 after the elements @c trace and @c host, as both the host
1457 and the trace definition must be known when @c trace_connect
1458 is parsed; the order of @c trace and @c host is arbitrary.
1461 #### @c trace attributes ####
1464 | Attribute name | Mandatory | Values | Description |
1465 | --------------- | --------- | ---------------------- | ----------- |
1466 | id | yes | String | Identifier of this trace; this is the name you pass on to @c trace_connect. |
1467 | 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. |
1468 | trace_periodicity | yes | String | This is the same as for @ref pf_tag_host "hosts" (see there for details) |
1470 Here is an example of trace when no file name is provided:
1473 <trace id="myTrace" periodicity="1.0">
1480 #### @c trace_connect attributes ####
1482 | Attribute name | Mandatory | Values | Description |
1483 | --------------- | --------- | ---------------------- | ----------- |
1484 | kind | no | HOST_AVAIL@|POWER@|<br/>LINK_AVAIL@|BANDWIDTH@|LATENCY (Default: HOST_AVAIL) | Describes the kind of trace. |
1485 | trace | yes | String | Identifier of the referenced trace (specified of the trace's @c id attribute) |
1486 | element | yes | String | The identifier of the referenced entity as given by its @c id attribute |
1488 @section pf_hints Hints, tips and frequently requested features
1490 Now you should know at least the syntax and be able to create a
1491 platform by your own. However, after having ourselves wrote some platforms, there
1492 are some best practices you should pay attention to in order to
1493 produce good platform and some choices you can make in order to have
1494 faster simulations. Here's some hints and tips, then.
1496 @subsection pf_hints_search Finding the platform example that you need
1498 Most platform files that we ship are in the @c examples/platforms
1499 folder. The good old @c grep tool can find the examples you need when
1500 wondering on a specific XML tag. Here is an example session searching
1501 for @ref pf_trace "trace_connect":
1504 % cd examples/platforms
1505 % grep -R -i -n --include="*.xml" "trace_connect" .
1506 ./two_hosts_platform_with_availability_included.xml:26:<trace_connect kind="SPEED" trace="A" element="Cpu A"/>
1507 ./two_hosts_platform_with_availability_included.xml:27:<trace_connect kind="HOST_AVAIL" trace="A_failure" element="Cpu A"/>
1508 ./two_hosts_platform_with_availability_included.xml:28:<trace_connect kind="SPEED" trace="B" element="Cpu B"/>
1509 ./two_hosts.xml:17: <trace_connect trace="Tremblay_power" element="Tremblay" kind="SPEED"/>
1512 @subsection pf_hint_generating How to generate different platform files?
1514 This is actually a good idea to search for a better platform file,
1515 that better fit the need of your study. To be honest, the provided
1516 examples are not representative of anything. They exemplify our XML
1517 syntax, but that's all. small_platform.xml for example was generated
1518 without much thought beyond that.
1520 The best thing to do when possible is to write your own platform file,
1521 that model the platform on which you run your code. For that, you
1522 could use <a href="https://gitlab.inria.fr/simgrid/platform-calibration">our
1523 calibration scripts</a>. This leads to very good fits between the
1524 platform, the model and the needs. The g5k.xml example resulted of
1525 such an effort, which also lead to <a href="https://github.com/lpouillo/topo5k/">an
1526 ongoing attempt</a> to automatically extract the SimGrid platform from
1527 the <a href="http://grid5000.fr/">Grid'5000</a> experimental platform.
1528 But it's hard to come up with generic models. Don't take these files
1529 too seriously. Actually, you should always challenge our models and
1530 your instanciation if the accuracy really matters to you (see <a
1531 href="https://hal.inria.fr/hal-00907887">this discussion</a>).
1533 But such advices only hold if you have a real platform and a real
1534 application at hand. It's moot for more abstract studies working on
1535 ideas and algorithms instead of technical artefacts. Well, in this
1536 case, there unfortunately is nothing better than this old and rusty
1537 <a href="http://pda.gforge.inria.fr/tools/download.html">simulacrum</a>.
1538 This project is dormant since over 10 years (and you will have to
1539 update the generated platforms with <tt>bin/simgrid_update_xml</tt> to
1540 use them), but that's the best we have for this right now....
1542 @subsection pf_zone_h Zone Hierarchy
1543 The network zone design allows SimGrid to go fast, because computing route is
1544 done only for the set of resources defined in the current zone. If you're using
1545 only a big zone containing all resource with no zone into it and you're
1546 using Full model, then ... you'll loose all interest into it. On the
1547 other hand, designing a binary tree of zone with, at the lower level,
1548 only one host, then you'll also loose all the good zone hierarchy can
1549 give you. Remind you should always be "reasonable" in your platform
1550 definition when choosing the hierarchy. A good choice if you try to
1551 describe a real life platform is to follow the zone described in
1552 reality, since this kind of trade-off works well for real life
1555 @subsection pf_exit_zone Exit Zone: why and how
1556 Users that have looked at some of our platforms may have notice a
1557 non-intuitive schema ... Something like that:
1561 <zone id="zone_4" routing="Full">
1562 <zone id="exitzone_4" routing="Full">
1563 <router id="router_4"/>
1565 <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"/>
1566 <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"/>
1567 <link id="4_1" bandwidth="2250000000" latency="5E-5"/>
1568 <link id="4_2" bandwidth="2250000000" latency="5E-5"/>
1569 <link id="bb_4" bandwidth="2250000000" latency="5E-4"/>
1570 <zoneroute src="cl_4_1"
1572 gw_src="c_4_1-cl_4_1_router"
1573 gw_dst="c_4_2-cl_4_2_router">
1574 <link_ctn id="4_1"/>
1575 <link_ctn id="bb_4"/>
1576 <link_ctn id="4_2"/>
1578 <zoneroute src="cl_4_1"
1580 gw_src="c_4_1-cl_4_1_router"
1582 <link_ctn id="4_1"/>
1583 <link_ctn id="bb_4"/>
1585 <zoneroute src="cl_4_2"
1587 gw_src="c_4_2-cl_4_2_router"
1589 <link_ctn id="4_2"/>
1590 <link_ctn id="bb_4"/>
1595 In the zone_4, you have an exitzone_4 defined, containing only one router,
1596 and routes defined to that zone from all other zone (as cluster is only a
1597 shortcut for an zone, see cluster description for details). If there was
1598 an upper zone, it would define routes to and from zone_4 with the gateway
1599 router_4. It's just because, as we did not allowed (for performances
1600 issues) to have routes from an zone to a single host/router, you have to
1601 enclose your gateway, when you have zone included in your zone, within an
1602 zone to define routes to it.
1604 @subsection pf_P2P_tags P2P or how to use coordinates
1605 SimGrid allows you to use some coordinated-based system, like vivaldi,
1606 to describe a platform. The main concept is that you have some peers
1607 that are located somewhere: this is the function of the
1608 <b>coordinates</b> of the @<peer@> or @<host@> tag. There's nothing
1609 complicated in using it, here is an example:
1612 <?xml version='1.0'?>
1613 <!DOCTYPE platform SYSTEM "https://simgrid.org/simgrid.dtd">
1614 <platform version="4">
1616 <zone id="zone0" routing="Vivaldi">
1617 <host id="100030591" coordinates="25.5 9.4 1.4" speed="1.5Gf" />
1618 <host id="100036570" coordinates="-12.7 -9.9 2.1" speed="7.3Gf" />
1620 <host id="100429957" coordinates="17.5 6.7 18.8" speed="8.3Gf" />
1625 Coordinates are then used to calculate latency (in microseconds)
1626 between two hosts by calculating the distance between the two hosts
1627 coordinates with the following formula: distance( (x1, y1, z1), (x2,
1628 y2, z2) ) = euclidian( (x1,y1), (x2,y2) ) + abs(z1) + abs(z2)
1630 In other words, we take the euclidian distance on the two first
1631 dimensions, and then add the absolute values found on the third
1632 dimension. This may seem strange, but it was found to allow better
1633 approximations of the latency matrices (see the paper describing
1636 Note that the previous example defines a routing directly between hosts but it could be also used to define a routing between zone.
1637 That is for example what is commonly done when using peers (see Section @ref pf_peer).
1639 <?xml version='1.0'?>
1640 <!DOCTYPE platform SYSTEM "https://simgrid.org/simgrid.dtd">
1641 <platform version="4">
1643 <zone id="zone0" routing="Vivaldi">
1644 <peer id="peer-0" coordinates="173.0 96.8 0.1" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us"/>
1645 <peer id="peer-1" coordinates="247.0 57.3 0.6" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1646 <peer id="peer-2" coordinates="243.4 58.8 1.4" speed="730Mf" bw_in="13.38MBps" bw_out="1.024MBps" lat="500us" />
1650 In such a case though, we connect the zone created by the <b>peer</b> tag with the Vivaldi routing mechanism.
1651 This means that to route between zone1 and zone2, it will use the coordinates of router_zone1 and router_zone2.
1652 This is currently a convention and we may offer to change this convention in the DTD later if needed.
1653 You may have noted that conveniently, a peer named FOO defines an zone named FOO and a router named router_FOO, which is why it works seamlessly with the <b>peer</b> tag.
1656 @subsection pf_routing_howto_choose_wisely Choosing wisely the routing model to use
1659 Choosing wisely the routing model to use can significantly fasten your
1660 simulation/save your time when writing the platform/save tremendous
1661 disk space. Here is the list of available model and their
1662 characteristics (lookup: time to resolve a route):
1664 @li <b>Full</b>: Full routing data (fast, large memory requirements,
1666 @li <b>Floyd</b>: Floyd routing data (slow initialization, fast
1667 lookup, lesser memory requirements, shortest path routing only).
1668 Calculates all routes at once at the beginning.
1669 @li <b>Dijkstra</b>: Dijkstra routing data (fast initialization, slow
1670 lookup, small memory requirements, shortest path routing only).
1671 Calculates a route when necessary.
1672 @li <b>DijkstraCache</b>: Dijkstra routing data (fast initialization,
1673 fast lookup, small memory requirements, shortest path routing
1674 only). Same as Dijkstra, except it handles a cache for latest used
1676 @li <b>None</b>: No routing (usable with Constant network only).
1677 Defines that there is no routes, so if you try to determine a
1678 route without constant network within this zone, SimGrid will raise
1680 @li <b>Vivaldi</b>: Vivaldi routing, so when you want to use coordinates
1681 @li <b>Cluster</b>: Cluster routing, specific to cluster tag, should
1684 @subsection pf_loopback I want to specify the characteristics of the loopback link!
1686 Each routing model automatically adds a loopback link for each declared host, i.e.,
1687 a network route from the host to itself, if no such route is declared in the XML
1688 file. This default link has a bandwidth of 498 Mb/s, a latency of 15 microseconds,
1689 and is <b>not</b> shared among network flows.
1691 If you want to specify the characteristics of the loopback link for a given host, you
1692 just have to specify a route from this host to itself with the desired characteristics
1693 in the XML file. This will prevent the routing model to add and use the default
1696 @subsection pf_switch I want to describe a switch but there is no switch tag!
1698 Actually we did not include switch tag. But when you're trying to
1699 simulate a switch, assuming
1700 fluid bandwidth models are used (which SimGrid uses by default unless
1701 ns-3 or constant network models are activated), the limiting factor is
1702 switch backplane bandwidth. So, essentially, at least from
1703 the simulation perspective, a switch is similar to a
1704 link: some device that is traversed by flows and with some latency and
1705 so,e maximum bandwidth. Thus, you can simply simulate a switch as a
1707 can be connected to this "switch", which is then included in routes just
1711 @subsection pf_multicabinets I want to describe multi-cabinets clusters!
1713 You have several possibilities, as usual when modeling things. If your
1714 cabinets are homogeneous and the intercabinet network negligible for
1715 your study, you should just create a larger cluster with all hosts at
1718 In the rare case where your hosts are not homogeneous between the
1719 cabinets, you can create your cluster completely manually. For that,
1720 create an As using the Cluster routing, and then use one
1721 <cabinet> for each cabinet. This cabinet tag can only be used an
1722 As using the Cluster routing schema, and creating
1724 Be warned that creating a cluster manually from the XML with
1725 <cabinet>, <backbone> and friends is rather tedious. The
1726 easiest way to retrieve some control of your model without diving into
1727 the <cluster> internals is certainly to create one separate
1728 <cluster> per cabinet and interconnect them together. This is
1729 what we did in the G5K example platform for the Graphen cluster.
1731 @subsection pf_platform_multipath I want to express multipath routing in platform files!
1733 It is unfortunately impossible to express the fact that there is more
1734 than one routing path between two given hosts. Let's consider the
1735 following platform file:
1738 <route src="A" dst="B">
1741 <route src="B" dst="C">
1744 <route src="A" dst="C">
1749 Although it is perfectly valid, it does not mean that data traveling
1750 from A to C can either go directly (using link 3) or through B (using
1751 links 1 and 2). It simply means that the routing on the graph is not
1752 trivial, and that data do not following the shortest path in number of
1753 hops on this graph. Another way to say it is that there is no implicit
1754 in these routing descriptions. The system will only use the routes you
1755 declare (such as <route src="A" dst="C"><link_ctn
1756 id="3"/></route>), without trying to build new routes by aggregating
1759 You are also free to declare platform where the routing is not
1760 symmetrical. For example, add the following to the previous file:
1763 <route src="C" dst="A">
1769 This makes sure that data from C to A go through B where data from A
1770 to C go directly. Don't worry about realism of such settings since
1771 we've seen ways more weird situation in real settings (in fact, that's
1772 the realism of very regular platforms which is questionable, but
1773 that's another story).
