JP2012147271A - Monitoring controller, communication route setting method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce power consumption by introducing sleep to line cards at both ends of a link.SOLUTION: The invention calculates a utilization rate of a link on a network from load information collected from the network; makes a calculation using Dijkstra's algorithm from a traffic matrix generated from flow information, the load information, and network configuration information when the maximum utilization rate oversteps the thresholds defining the range between upper and lower limits; sequentially extracts links to which sleep can be introduced until the utilization rate of the link having the largest load oversteps the threshold; and calculates a set of weighting values of the links so that loads are maximally distributed in an OSPF protocol for topology performing sleep to the extracted links. When the threshold is overstepped, the invention acquires link topology before overstepping the threshold and calculates a link cost, determines final link cost and a route based on the link cost, and sets the link cost to a real network and introduces the sleep to links that are not used.

Description

  The present invention relates to a monitoring control apparatus, a communication path setting method, and a program, and more particularly, to a network apparatus, a communication path setting method, and a program for reducing power consumption of a router in a managed IP (Internet Protocol) relay network. About.

  In routers, unlike servers, power consumption in standby does not differ significantly from maximum load (about 7% difference), so use the sleep state to greatly reduce standby power. Has been presented (see, for example, Non-Patent Document 1).

  A technique has also been developed in which traffic flowing in a specific router is distributed to other routers by performing communication path control in units of networks and sleep is introduced (for example, see Non-Patent Document 2).

  In addition, for each router, Rate Adaptation (see, for example, Non-Patent Document 3) that reduces standby power by introducing a small-scale sleep during node operation using a line card buffer is also attracting attention.

Joseph Chabarek, Joel Sommers, Paul Barford, et.al "Power Awareness in Network Design and Routing" In Proc. IEEE INFOCOM, 2008. http://pages.cs.wisc.edu/~pb/infocom08a_final.pdf Hirohito Nomura, Ken Yada, Hiroshi Yamada, "A Study on Network Power Saving Considering Device Power Characteristics", IEICE, Society Conference, September 2010. Serigiu Nedvschi, Lucian Popa, et. Al, "Reducing Network Energy consumption via Rate-Adaptation and Sleeping", 2008 http://www.eecs.berkeley.edu/Pubs/TechRps/2007/EECS-2007-128.pdf

  However, most of the currently proposed methods for reducing the power consumption of routers require the introduction of routers equipped with MPLS (Multi-Protocol Label Switching) and Route Adaptation (Non-patent Document 3). ing.

  For this reason, in a current general network that does not have a router that can use MPLS or that uses OSPF (Open Shortest Path First) for routing, a router that has entered a sleep state is determined to be powered off and is deleted from the path. Therefore, it is not possible to use a technology for easily introducing sleep of these routers.

  Although a method adapted to OSPF routing has also been proposed (Non-patent Document 2), it is an algorithm that uses the maximum link utilization rate as the termination condition. There is a problem that the introduction ends even if there is room.

  The present invention has been made in view of the above points, and in a communication network in which a communication path in the network is controlled by an OSPF routing protocol, it is possible to reduce power consumption by introducing sleep to the line cards at both ends of the link. An object of the present invention is to provide a possible network device, a communication path setting method, and a program.

In order to solve the above problems, the present invention (Claim 1) is a monitoring control device for performing power saving in a communication network in which a communication path in the network is controlled by an OSPF routing protocol,
Storage means for storing network configuration information, network topology, traffic flow information, interface load information;
Communication interface means for performing communication with the network, acquiring the traffic flow information and interface load information and storing the information in the storage means;
Traffic matrix generation means for generating a traffic matrix from the traffic flow information;
Link utilization rate calculating means for obtaining a link utilization rate based on the traffic matrix from the interface load information;
When the link utilization rate exceeds the predetermined upper and lower thresholds, the sleep target link that does not flow traffic is excluded from the network topology of the storage means by changing the weight value of each link. Sleep introduction calculation means to create a new topology excluding the link,
A combination calculation means for calculating a combination of link weight values for leveling traffic with respect to the new topology created by the sleep introduction calculation means;
The sleep introduction calculation means and the combination calculation means are performed until the maximum link utilization rate exceeds a predetermined threshold value, and when the threshold value is exceeded, the topology immediately before the threshold value is obtained, and the topology is obtained. Load distribution means for calculating a link cost and calculating a combination of link weight values such that the load is most distributed in the OSPF routing protocol from the link cost;
Applying / introducing means for applying a topology excluding a link to be sleep to the network based on the combination of the weight values and link information capable of sleeping.

  As described above, the present invention does not use a technique such as MPLS or Rate Adaptation, which has devices that cannot be implemented at present, and in the general OSPF protocol network, the OSPF link set in each router. Only by changing the weight value (link weight), it is possible to reset the communication path in which sleep is introduced, and to realize power saving of the router.

  In addition, in a network consisting of links that do not introduce sleep, load distribution is performed only by manipulating the weight value of the OSPF link, thereby reducing the high load on a specific link and further introduction of sleep can be expected. It becomes like this.

It is a network block diagram in one embodiment of this invention. It is a block diagram of the monitoring control apparatus in one embodiment of this invention. It is a block diagram of the calculating part in one embodiment of this invention. It is a figure which shows the outline | summary operation | movement of the whole network in one embodiment of this invention. It is a flowchart of the sleep introduction calculation process in one embodiment of the present invention. It is a flowchart of the load distribution calculation process in one embodiment of this invention. It is an example of the values of φ at c _a = 1 obtained by the arithmetic unit according to an embodiment of the present invention _{(l a, c a a)} . It is a flowchart of the whole operation | movement of the monitoring control apparatus in one embodiment of this invention.

  Embodiments of the present invention will be described below with reference to the drawings.

  First, terms used in this specification will be described.

  Weight value: The link weight used for OSPF route calculation set by the network administrator for each individual router.

  Route cost: the sum of the link weight values on the shortest path between two nodes calculated by the Dijkstra method.

  Cost: The value of the cost function used in load balancing calculations.

  Next, the preconditions of the present invention will be described.

  1. About half of the power consumption of the router is constituted by the chassis, and most of the rest is constituted by the line cards. The power consumption of the line card increases as the processing speed increases (see Non-Patent Document 1).

  2. In each router constituting the network, sleep can be introduced for each component. In the sleep state, power consumption is almost zero, but processing cannot be performed.

  3. By using traffic flow measurement technology such as NetFlow and sFlow, it is possible to collect information on the flow that flows through each device to a specific device (flow collector) (for example, “Reference 1:“ Cisco NetFlow v3 ”http: / /www.cisco.com/warp/public/cc/pd/nemnsw/nesv/prodlit/2033_pp.pdf "," Reference 2: sFlow http://www.sflow.org/about/index.php ").

  4). 3. above. It is possible to estimate the traffic matrix from the flow information collected in step 3 using techniques such as Document 3 (Yoshiaki Tanaka, Daisuke Tachibana, Masato Uchida et al. .

  5). A network device can notify load information of its interface to a specific monitoring device via SNMP or the like.

  First, the network configuration of the present invention will be described.

  FIG. 1 shows a network configuration according to an embodiment of the present invention.

  The network 1 targeted by the present invention is a single AS (Autonomous System) using OSPF as a routing protocol, and has a redundant structure. The traffic flowing through the network 1 varies greatly for each ground and the entire flow rate during the day and night, weekdays or holidays.

  The network 1 includes a plurality of AS boundary routers 10, and it is assumed that standby power is large in the line cards of these routers 10. Each router 10 notifies the monitoring control device 20 of traffic flow information flowing through itself.

  The monitoring control device 20 is a device that is connected to the network 1 and manages the configuration information of the network 1. The monitoring control device 20 functions as a flow collector, acquires flow information and load information from each router 10, and routes routes according to traffic conditions. Calculate. Further, each router 10 can be accessed by Telnet or SSH (Secure SHell) to change the setting.

  FIG. 2 shows a configuration of the monitoring control apparatus according to the embodiment of the present invention. The monitoring control device 20 includes a communication interface unit 21, a storage unit 22, and a calculation unit 23.

  The communication interface unit 21 is connected to the network 1, has an IP address, and performs communication processing with each managed device.

  The storage unit 22 stores network device setting information, network topology before sleep introduction, after sleep introduction, exceeding a threshold, after network introduction, immediately before the threshold, flow matrix, load information of each interface, and the like.

  The calculation unit 23 acquires information necessary for calculation and stores calculation results and device information through calculations for periodic network device management and sleep introduction, and exchange with the storage unit 22. Specifically, periodic information acquisition processing and power saving processing are performed.

  Periodic information acquisition processing includes processing for periodically acquiring load information and setting information of each device in the network by SNMP and the like, and acquiring flow information from the router 10 by NetFlow or the like, and storing it in the storage unit 22. Store.

  As power saving processing, estimation of traffic matrix from flow information, calculation of sleep possible link, calculation of weight value combination to distribute load by load distribution calculation, combination processing of sleep introduction calculation and load distribution calculation, Apply the settings to the actual network and perform sleep introduction processing.

  FIG. 3 shows the configuration of the calculation unit in one embodiment of the present invention.

  The calculation unit 23 uses the information management unit 231 that performs the above-described periodic information acquisition processing, the link usage rate calculation unit 232 that calculates the link usage rate (load), and the flow information acquired from the storage unit 22 described above in Reference 3. A traffic matrix generation unit 233 that generates a traffic matrix by the method shown in FIG. 5, a sleep introduction calculation unit 234 that performs sleep introduction calculation, a load distribution unit 235 that performs calculation to distribute a load based on the link utilization rate, An application / introduction unit 236 that introduces sleep and a memory 237 that stores calculation results calculated in the calculation unit 23.

  Next, an outline of the operation in the above configuration will be described.

  FIG. 4 shows an outline operation of the entire network in the embodiment of the present invention.

  Step 0) In an initial state, the network 1 determines a route according to a routing protocol set in advance.

  Step 1) Each router 10 notifies the monitoring control device 20 of the load information applied to its own interface by SNMP. Further, the edge router collects flow information necessary for creating a traffic matrix and transmits the flow information to the monitoring controller 20 at regular intervals.

  Step 2) The communication interface unit 21 of the monitoring control device 20 acquires the load information and flow information notified from each router 10 and stores them in the storage unit 22. The link usage rate calculation unit 232 of the calculation unit 23 acquires load information from the storage unit 22 and obtains the usage rate of the link on the network 1. When the maximum utilization rate exceeds a range (between the upper limit value and the lower limit value) determined by a predetermined threshold, the calculation unit 23 performs route determination considering power efficiency by the following operation. Details will be described later.

  Step 3) The computing unit 23 reads the flow information from the storage unit 22 and creates a traffic matrix in the traffic matrix creating unit 233.

  Step 4) In the sleep introduction calculation unit 234, the calculation unit 23 performs a calculation using the Dijkstra method from the traffic matrix and the load information generated in Step 3 and the network configuration information, and the utilization rate of the link with the largest load is a threshold value. The links where sleep can be introduced are selected until θ is exceeded. Details will be described later.

  Step 5) The calculation unit 23 calculates, in the load distribution unit 235, a combination of link weight values such that the load is most distributed in the OSPF protocol with respect to the topology in which the link calculated in Step 4 is sleeping. To do.

  Step 6) While the maximum link utilization rate falls below the threshold θ by the calculation in Step 5, the above Steps 4 and 5 are repeated.

  Step 7) When the threshold value θ exceeds the threshold θ, the load distribution unit 235 obtains the link topology immediately before exceeding the threshold value θ in Step 4 and uses the load distribution in Step 5 as the final link cost.

  Step 8) When the final link cost and the route determination based on the final link cost are determined, the link cost is set to the actual network 1 and the sleep is introduced to the unused link via the communication interface unit 21 of the monitoring control device 20. . Details will be described later.

  Next, the sleep introduction calculation process in step 4 will be described.

  FIG. 5 is a flowchart of the sleep introduction calculation process according to the embodiment of the present invention.

  Prior to the sleep introduction calculation process, the traffic matrix creation unit 233 of the computation unit 23 creates a matrix of traffic flowing through the network 1 based on the traffic flow information stored in the storage unit 22 via the communication interface unit 21. And stored in the memory 237 inside the calculation unit 23. As a method for creating the traffic matrix, for example, a method according to Reference 4 “sFlow http://www.sflow.org/about/index.php” is used. At this time, in order to ensure the communication between all the edges, it is assumed that traffic having a value lower than the minimum traffic amount is given to the edge-to-ground where no flow is observed.

  Further, the communication interface unit 21 stores the load information of all the links of the network acquired by SNMP in the storage unit 22, and the calculation unit 23 sets the usage rate to the threshold θ based on the load information of all the links. Determine if there are more links. If there is no link exceeding the threshold θ, the process shown in FIG. 5 is performed. If there is a link exceeding the threshold θ, the process shown in FIG. 5 is not performed, and the process proceeds to the load distribution process in step 5.

  Step 401) The information management unit 231 of the calculation unit 23 obtains the traffic matrix generated in Step 3, the network topology including the bandwidth and the weight value, and the load information of each link from the storage unit 22.

  Step 402) The sleep introduction calculation unit 234 sets the current link weight value and the threshold value when there is a link having a utilization rate exceeding a preset threshold value θ or when verification of the sleep introduction of all links is completed. The link weight value setting stored in the memory 237 in the previous processing of Steps 403 to 408 having the link weight value setting not exceeding the value is read and output to the load distribution unit 235. If there is no link exceeding the threshold θ, the process proceeds to step 403.

  Step 403) The sleep introduction calculation unit 234 stores the current weight values of all links in the memory 237.

  Step 404) In order to remove a link having a low utilization rate but no alternative route from the sleep introduction target, a link information list excluding links to which a sleep flag described later is set is created and stored in the internal memory 237.

  Step 405) The sleep introduction calculation unit 234 identifies a link having a minimum utilization rate that is less than the threshold θ in the created link list, and sets the sleep flag “1”.

  Step 406) Further, the cost of the link with the sleep flag set is increased, and traffic is prevented from flowing. For example, the set cost is set to the sum of the weights of all links + 1.

  Step 407) Based on the traffic matrix in the memory 237 and the network configuration information read from the storage unit 22, the calculation based on the Dijkstra method is performed, and the utilization rate of each link is calculated and stored in the memory 237.

  Step 408) The sleep introduction calculation unit 234 first determines whether there is a link having a utilization rate exceeding the threshold θ with respect to the utilization rate of all links calculated in Step 407, and the link having the utilization rate exceeding the threshold θ. If there is a network topology, the network topology (that exceeds / does not exceed the threshold) when sleep is introduced is output to the load distribution unit 235. On the other hand, if there is no utilization rate link exceeding the threshold θ, the process proceeds to step 409.

  Step 409) It is determined whether the utilization rate of the link for which the sleep flag was set immediately before is 0. If the utilization rate is 0, the process proceeds to step 410. If the utilization rate is greater than 0, sleep is introduced to the link. As a result, a ground-to-ground connection is lost, and the process proceeds to step 402 without introducing sleep.

  Step 410) If the utilization rate is 0 in Step 409, the link is removed from the network, and the process proceeds to Step 402, where the utilization rate exceeds the threshold θ or all links are registered in the sleep flag list. The above processing is repeated until an information list cannot be created.

  Next, the load distribution calculation process in step 5 will be described.

The load distribution unit 235 of the calculation unit 23 performs the load distribution calculation in Step 5 as document 5 “Fortz, B. Thorup, M.“ Internet traffic engineering by optimizing OSPF weights ”, Proc. 19 ^th IEEE Conf. Computer Communications (INFOCOM ), pp. 519-528 http://ieee.org/xpls/abs_all.jsp?arnumber=832225&tag=1 ” A combination of link weight values for equalizing the utilization rate is calculated. This load leveling uses the property of OSPF that traffic is distributed when there are multiple paths with the same path cost between certain nodes, so that the slope of the cost function varies according to the threshold θ. Changes have been made.

  FIG. 6 is a flowchart of the load distribution calculation process according to the embodiment of the present invention.

In the load distribution calculation by the load distribution unit 235 shown below, the cost function shown below is used which takes the shape shown in FIG. 7 so that the utilization rate of each link is less than the threshold θ. The cost Φ of the network is obtained by the following formula as a cost function φ (l _a , c _a ), a set A of the link a of the network, a load l _{a of} the link a, and a capacity c _a of the link a.

ステップ５０１） 前述のスリープ導入計算部２３４から出力されたネットワークトポロジとメモリ２３７からトラヒック行列、及び閾値θを取得し、上記のコスト関数のパラメータを決定する。

Step 501) The network topology output from the sleep introduction calculation unit 234 and the traffic matrix and the threshold value θ are acquired from the memory 237, and the parameters of the cost function are determined.

Step 502) The load distribution unit 235 obtains a random value from W = {1,..., W _max } as an initial solution and sets it as a weight value for each link (a combination of weight values for this link is set as an initial solution). And).

  Step 503) The load distribution unit 235 calculates a Shortest Path.

  Step 504) The link usage rate calculation unit 232 calculates the usage rate of each link using the traffic matrix, the initial solution, and the network topology.

  Step 505) The load distribution unit 235 applies the usage rate calculated by the link usage rate calculation unit 232 to the above cost function to calculate the cost of the initial solution. The cost function has a property that the cost increases as the link utilization rate increases.

  Step 506) The load distribution unit 235 stores the current solution and cost in the hash table 1 (HT1) on the memory 237, and uses this as a tabu list.

Step 507) N _n neighboring solutions w ′ are created from the current solution stored in HT1 according to the technique of the above-mentioned document 5.

  Step 508) It is determined whether the obtained neighborhood solution w ′ is included in HT1. If included, the process proceeds to Step 509. If not included, the process proceeds to Step 510.

  Step 509) The neighborhood solution w ′ is discarded, and the process proceeds to Step 507.

Step 510) It is determined whether or not the neighborhood solution w ′ has reached N _n . If reached, the process proceeds to Step 511, and if not, the process proceeds to Step 507.

  Step 511) Extract neighboring solutions that are not included in the hash table 2 (HT2) storing the cost-evaluated solution and the cost value.

  Step 512) The Shortest Path of the neighborhood solution extracted in Step 511 is calculated.

  Step 513) The link usage rate calculation unit 232 calculates the load of the extracted solution.

  Step 514) The load distribution unit 235 calculates the cost of the extracted solution. The calculation method is the same as in step 505.

  Step 515) The extracted solution and its cost are stored in HT2 on the memory 237.

  Step 516) The load distribution unit 235 extracts the nearest solution having the lowest cost as the next solution.

  Step 517) The extracted lowest cost solution and cost value are stored in HT1.

Step 518) HT1 number of elements in it is determined whether reaches _{N i,} if not reached then proceeds to step 507, if it reaches the process proceeds to step 519.

Step 519) The process of steps 507-518 repeated _{N i} times, and outputs the application-introducing portion 236 most cost a was low solution as the optimal solution (the combination weight values) in HT1. It should be noted that at a certain number of repetitions, a part of the solution elements are perturbed to form a new solution, and the search range is changed.

  Next, the overall operation in the monitoring control device 23 will be described.

  FIG. 8 is a flowchart of the overall operation of the monitoring control apparatus according to the embodiment of the present invention. As an assumption of the following processing, the information management unit 231 of the calculation unit 23 always collects load information of each device of the network 1 via the communication interface 21 at a constant time interval and stores it in the storage unit 22. And

  Step 701) The link usage rate calculation unit 232 of the calculation unit 23 calculates the calculated load information of the interface of each device of the network 1, and the usage rate of the link with the highest usage rate is equal to or higher than a predetermined upper threshold α. If it is less than the lower threshold β, the process proceeds to step 702, and if β is less than α, the process waits without doing anything.

  Step 702) The traffic matrix creation unit 233 of the computation unit 23 reads out the flow information obtained from the router 10 of the network 1 from the storage unit 22, creates a traffic matrix, and stores it in the memory 237.

  Step 703) The sleep introduction calculation unit 234 of the calculation unit 23 reads the traffic matrix from the memory 237 and the topology information from the storage unit 22, performs a sleep introduction process, and identifies a link where sleep introduction is possible. At this time, the topology after sleep introduction is stored under the following two conditions.

Condition a: Topology immediately after the maximum link utilization exceeds the threshold value θ;
Condition b: Topology immediately before the maximum link utilization rate exceeds the threshold θ;
The above condition a is used to introduce a further sleep by causing the maximum link utilization rate to decrease due to the load distribution process. Condition b is used when the maximum link utilization rate after suppression by load distribution processing exceeds θ.

  Step 704) The sleep introduction calculation unit 234 determines whether or not the maximum link utilization rate after the distributed processing exceeds the threshold θ, and if it exceeds (over flag = 1), the process proceeds to Step 706. Otherwise, the process proceeds to step 705.

  Step 705) Select condition a as the input topology, and proceed to Step 707.

  Step 706) The condition b is selected as the input topology, and the process proceeds to Step 707.

  Step 707) The load distribution unit 235 calculates a combination of weight values so that the link utilization rate is leveled in the topology after sleep introduction.

  Step 708) The link usage rate calculation unit 232 calculates the usage rate of each link, and the load distribution unit 235 acquires the usage rate of the link and determines whether the maximum link usage rate exceeds the threshold θ. . If so, the process proceeds to step 709; otherwise, the process proceeds to step 710.

  Step 709) If the maximum link utilization rate> θ, over_flag = 1 is set, the process proceeds to Step 704, and the load distribution calculation process in the topology under the condition b is started (Steps 706, 707).

  Step 710) If the maximum link utilization rate ≦ θ, it is determined whether over_flag = 1. If over_flag = 1, the process proceeds to step 711. If over_flag = 0 (initial value), the process proceeds to step 703. The sleep introduction process is performed using the topology with the output weight value as input.

  Step 711) The load distribution unit 235 sets the sleep link weight value. At this time, the weight value of the link treated as sleep is set to a value larger than the sum of the other links.

  Step 712) The application / introduction unit 236 accesses each router 10 of the actual network 1 using Telnet or SSH via the communication interface 21, and sets a final link weight value.

  Step 713) The application / introduction unit 236 sleeps the interfaces at both ends when the traffic does not flow on the sleep target link due to the route reset, and reduces power consumption.

  The present invention constructs the operation of the constituent elements of the monitoring control device 20 shown in FIGS. 2 and 3 in the above-described embodiment as a program, and installs and executes it on a computer used as the monitoring control device, or a network It is possible to circulate through.

  The present invention is not limited to the above-described embodiment, and various modifications and applications can be made within the scope of the claims.

DESCRIPTION OF SYMBOLS 1 Network 10 AS border router 20 Monitoring control apparatus 21 Communication interface part 22 Storage part 23 Calculation part 231 Information management part 232 Link utilization rate calculation part 233 Traffic matrix preparation part 234 Sleep introduction calculation part 235 Load distribution part 236 Application | installation / introduction part

Claims (9)

A monitoring and control device for power saving in a communication network in which the communication path in the network is controlled by the OSPF (Open shortest Path First) routing protocol,
Storage means for storing network configuration information, network topology, traffic flow information, interface load information;
Communication interface means for performing communication with the network, acquiring the traffic flow information and interface load information and storing the information in the storage means;
Traffic matrix generation means for generating a traffic matrix from the traffic flow information;
Link utilization rate calculating means for obtaining a link utilization rate based on the traffic matrix from the interface load information;
When the link utilization rate exceeds the predetermined upper and lower thresholds, the sleep target link that does not flow traffic is excluded from the network topology of the storage means by changing the weight value of each link. Sleep introduction calculation means to create a new topology excluding the link,
A combination calculation means for calculating a combination of link weight values for leveling traffic with respect to the new topology created by the sleep introduction calculation means;
The sleep introduction calculation means and the combination calculation means are performed until the maximum link utilization rate exceeds a predetermined threshold value, and when the threshold value is exceeded, the topology immediately before the threshold value is obtained, and the topology is obtained. Load distribution means for calculating a link cost and calculating a combination of link weight values such that the load is most distributed in the OSPF routing protocol from the link cost;
Applying / introducing means for applying a topology excluding a link to be sleep to the network based on the combination of the weight values and link information capable of sleeping;
A monitoring control apparatus comprising: The sleep introduction calculation means includes:
A link list in which links that have a link utilization rate lower than a predetermined threshold value but no alternative route are excluded from the sleep introduction targets is generated from all links, and links that are less than the predetermined threshold value and have the lowest utilization rate in the link list A first means for increasing the weight value of
Based on the traffic matrix and the network configuration information, each link usage rate is obtained using dynamic programming. If there is no link with a usage rate exceeding a predetermined threshold, is the usage rate of the previous link 0? If the link is 0, the link is removed from the link list, and if the usage rate is greater than 0, the link is not a sleep target. If there is a link with a usage rate exceeding the predetermined threshold, the link Second means for outputting the current link weight value or the immediately preceding weight value to the combination calculating means;
The monitoring control device according to claim 1, comprising: The second means includes
If the maximum link utilization rate exceeds the predetermined threshold, the link weight value immediately before the maximum link utilization rate exceeds the threshold value. Otherwise, the link immediately after the maximum link utilization rate exceeds the threshold value. The monitoring control apparatus according to claim 2, wherein a weight value is output to the combination calculation unit. The load balancing means includes
When the maximum link utilization rate of the combination of the link weight values exceeds a predetermined threshold, the link topology immediately before exceeding the threshold is acquired, and a cost function that increases the cost as the link utilization rate increases is used. The monitoring control apparatus according to claim 1, further comprising means for calculating a final link cost and obtaining a weight value of a link having the lowest link cost. A monitoring control method for power saving in a communication network in which a communication path in the network is controlled by the OSPF routing protocol,
Store network configuration information and network topology in storage means in the device,
A communication interface means for performing communication with the network, acquiring the traffic flow information and interface load information from the storage means and storing them in the storage means;
A traffic matrix generating means for generating a traffic matrix from the traffic flow information;
A link utilization rate calculating means for obtaining a link utilization rate based on the traffic matrix from the interface load information;
When the link utilization rate exceeds a predetermined upper limit and lower limit threshold, the sleep introduction calculation means changes the weight value of each link to change the sleep target link that does not flow traffic from the network topology of the storage means. A sleep introduction calculation step of creating a new topology excluding the sleep target link,
A load calculating means for calculating a combination of link weight values for leveling traffic with respect to the new topology created in the sleep introduction calculating step;
The load distribution means performs the sleep introduction calculation step and the combination calculation step until the maximum link utilization rate exceeds a predetermined threshold value, and if it exceeds the threshold value, acquires the topology immediately before the threshold value is exceeded. A load distribution step of calculating a link cost for the topology and calculating a combination of link weight values such that the load is most distributed in the OSPF routing protocol from the link cost;
An application / introduction step in which the application / introduction means applies a topology excluding the link to be sleep-based to the network based on the combination of the weight values and the sleepable link information;
A monitoring control method characterized by: In the sleep introduction calculation step,
A link list in which links that have a link utilization rate lower than a predetermined threshold value but no alternative route are excluded from the sleep introduction targets is generated from all links, and links that are less than the predetermined threshold value and have the lowest utilization rate in the link list A first step of increasing the weight value of
Based on the traffic matrix and the network configuration information, each link usage rate is obtained using dynamic programming. If there is no link with a usage rate exceeding a predetermined threshold, is the usage rate of the previous link 0? If the link is 0, the link is removed from the link list, and if the usage rate is greater than 0, the link is not a sleep target. If there is a link with a usage rate exceeding the predetermined threshold, the link A second step of outputting the current link weight value or the previous weight value of
The monitoring control method according to claim 5 comprising: In the second step,
If the maximum link utilization rate exceeds the predetermined threshold, the link weight value immediately before the maximum link utilization rate exceeds the threshold value. Otherwise, the link immediately after the maximum link utilization rate exceeds the threshold value. The monitoring control method according to claim 6, wherein a weight value is output. In the load balancing step,
When the maximum link utilization rate of the combination of the link weight values exceeds a predetermined threshold, the link topology immediately before exceeding the threshold is acquired, and a cost function that increases the cost as the link utilization rate increases is used. 6. The monitoring control method according to claim 5, wherein a final link cost is calculated and a weight value of a link having the lowest link cost is obtained.   The monitoring control program for functioning a computer as each means which comprises the monitoring control apparatus of any one of Claims 1 thru | or 4.

Images