CN104270291B - CDN network quality control method - Google Patents
CDN network quality control method Download PDFInfo
- Publication number
- CN104270291B CN104270291B CN201410567526.8A CN201410567526A CN104270291B CN 104270291 B CN104270291 B CN 104270291B CN 201410567526 A CN201410567526 A CN 201410567526A CN 104270291 B CN104270291 B CN 104270291B
- Authority
- CN
- China
- Prior art keywords
- node
- data center
- test
- interchanger
- network quality
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The present invention relates to a kind of content distributing network quality control method, content distributing network includes data center, bandwidth provider interchanger and fringe node, method and comprises the following steps:Monitoring equipment inside data center carries out ping testing to the node server of data center;The bandwidth provider interchanger networked from the monitoring equipment pair inside data center with data center carries out ping testing;Fringe node is tested from the monitoring equipment inside data center, if fringe node is controllable node, at least part of ping testing, downloading rate test and page Poll tests is carried out, if fringe node is uncontrollable node, carries out ping testing;Wherein for controllable node for that by network entry and can dispose the fringe node of the HTTP website services of downloading rate test and page Poll tests by the supervisor of data center, uncontrollable node is fringe node that cannot be by the supervisor of data center by network entry or the HTTP website services that cannot dispose downloading rate test and page Poll tests.
Description
Technical field
The present invention relates to CDN network quality test, particularly with regard to the method for CDN network quality monitoring.
Background technology
The full name of CDN is Content Delivery Network, i.e. content distributing network.The purpose is to by existing
Internet in increase by one layer of new network architecture, the content of website is published to the network " edge " closest to user, is made
User can obtain required content nearby, solve the crowded situation of Internet network, improve the response that user accesses website
Speed.Technically comprehensively solve is since network bandwidth is small, user's visit capacity is big, the use caused by the unequal reason of network point distribution
Family accesses the slow-footed problem of websites response.CDN is to establish and be covered on bearer network, the service by being distributed in different zones
The distributed network of node composition.Source contents are transferred to edge closest to user by it by certain rule, allow user
Required content is obtained nearby, reduces the bandwidth requirement to backbone network, improves the response speed that user accesses.From CDN technologies
Description is not difficult to find out, the information at remote data center can be transferred on home server by CDN technologies, and local user visits
Ask that home server can finishing service.Briefly, content issue net (CDN) is an overall system through tactic deployment
System, including 4 important documents of the redirection of distributed storage, load balancing, network request and Content Management, and Content Management and the overall situation
Network flow management (Traffic Management) be CDN core where.Pass through user's close and server load
Judgement, CDN ensure content by it is a kind of it is extremely efficient in a manner of for the request of user provide service.Generally speaking, content service
Based on cache server, also referred to as proxy caching, it be located at the edge of network, away from user only have " jump " (Single Hop) it
It is distant.Meanwhile proxy caching be one of content supplier's source server (data center for being usually located at CDN service provider) thoroughly
Bright mirror picture.Such framework enables CDN service provider to represent their clients, i.e. content provider, is carried to end user
For experience as well as possible, and these users can't stand request response time any delay.According to statistics, using CDN
Technology, can handle 70%~95% access to content amount of whole Website page, mitigate the pressure of server, improve website
Performance and scalability.
Typical CDN is made of five parts, mainly includes five parts:Source server, node server, GSLB controls
Server, Network Management System, Content Management System.Source server is exactly the data center server where real website.
Node server is exactly the server positioned at each cloud point, the web site contents that in store people frequently access on node server
Copy.GSLB control servers are dedicated with the highly reliable and equipment of high redundancy, its effect is by source server
Flow is transferred on node server working properly.GSLB can effectively ensure to access to need with the source of automatic decision user
Ask be assigned to from its in logic recently, on most healthy node server.Content Management System is provided according to content supplier
Content and target customers, requirements for access is made in advance judge whether by the caching of current content or caching how long or
It is not kept in node server and directly returns source server etc..
In order to ensure the normal operation of whole system, each CDN network has a Network Management System.It is responsible for realization
Distribution detection and alarm is concentrated, ensure to find failure, positioning failure within first time, and while network-in-dialing is monitored pair
The respective services of application layer are monitored in real time.Therefore, network quality test has the function that important that it is in CDN operations
Judge CDN new nodes network quality (such as decision node server or data center are with the presence or absence of failure etc.), meshed network is answered
Which region of the covering, and the important tool of monitoring meshed network quality.It is serviced at other for network quality at the same time
Play the role of important.
Existing network quality test is largely oneself controllable IP storehouse of test, main following several test methods:
1.PING is tested:
PING (Packet Internet Grope), the Internet packets survey meter, is a kind of to be used for test network connection amount
Program.PING sends an ICMP echo requests message to destination and reports whether receive desirable ICMP Echo Replies.
The excessive time that PING applications return cannot can effectively transmit the dissemination system of information slowly for excluding those all toos.
This test is used for reducing the number of distribution website test by terminal.The shortcomings that ping testing, is that ping testing can only be tested
Delay and packet drop end to end, but it is corresponding often to establish tcp also comprising DNS is accessed during user to access pages
And the time of file is pulled, although ping testing a degree of can react link state end to end, this test
It can not test out and all cover the time.
2. downloading rate is tested
Download refers between computer by way of http or File Transfer Protocol go to exchange data, the transmission between them
Speed is that most directly most intuitively a kind of test mode, this test mode are also network quality situation end to end for measurement one
Most common a kind of mode in CDN network test.Due to it is this test need to rely on tested end be it is controllable and
The server end that itself stablizes, can be uncontrollable so being just highly dependent on tested end.For example, if tested object is one
Common computer, and without externally authority of the offer Telnet and software download etc. for network test, then with regard to nothing
Method is downloaded rate test to the object.
3. page poll is tested
The opening speed of website just refers to the total time shown completely to Website page from click website, and page Poll is surveyed
Examination, which is exactly analog subscriber, asks to the test for eventually receiving pagefile content and being taken time, this survey from sending accession page
Examination is the test of most proper user terminal, during which the time spent in have search the DNS times, establish connection TCP need time, data to connect
Between time receiving and present the time.However, similar with downloading rate test mode, page Poll tests are also to rely on tested end
It is the server end that can be controlled and stablize in itself, can be uncontrollable so being just highly dependent on tested end.
Generally because the IP storehouses of coverage test are more at random, means of testing is single and depends on for existing network quality test
Oneself controllable test object IP address (such as downloading rate test and page poll tests just need the tested object to be
Controllable), it can not be unfolded accurately to test if a strange overlay area without the server of oneself.Test at the same time
Be all external client IP address, therefore closer to this end of CDN node network condition exist monitoring blind spot.So cause
Test coverage face is limited, test accuracy not high situation.
In addition, the test of existing CDN network can only all be immediately seen final test result, such as device A is to equipment B
The time spent in delay of test and packet loss, downloading rate and the poll pages.In fact, often through excessive between A to B
A interchanger or a plurality of network path, merely from the point of view of the test result from A to B, the communication that can only understand A to B goes wrong,
But can not decision problem which specific link appeared in.
The content of the invention
The technical problems to be solved by the invention are to provide a kind of CDN network quality control method, it can provide layering
Monitoring capacity.
The present invention is to propose a kind of content distributing network quality prison to solve the technical solution that above-mentioned technical problem uses
Prosecutor method, the content distributing network include data center, bandwidth provider interchanger and fringe node, and this method includes following
Step:Monitoring equipment inside the data center carries out PING (Packet Internet to the node server of the data center
Grope) test;The bandwidth provider interchanger networked from the monitoring equipment pair inside the data center with the data center carries out
Ping testing;Fringe node is tested from the monitoring equipment inside the data center, if wherein the fringe node is controllable
Node, then carry out at least part of ping testing, downloading rate test and page Poll tests, if the fringe node is can not
Node is controlled, then carries out ping testing;Wherein controllable node is that can pass through network entry and portion by the supervisor of data center
The fringe node of the HTTP website services of downloading rate test and page Poll tests is affixed one's name to, uncontrollable node is can not be by data
The supervisor at center is by network entry or cannot dispose the HTTP website services that downloading rate is tested and page Poll is tested
Fringe node.
In one embodiment of this invention, after to the server progress ping testing of the data center, further including to test
The delay of acquisition and packet loss are compared with one group of first reference value, to determine tested node server or node server
Whether the interchanger connected is normal.
In one embodiment of this invention, pair bandwidth provider interchanger networked with data center progress ping testing
Afterwards, the packet loss obtained will be tested compared with one second reference value by further including, whether just to determine the bandwidth provider interchanger
Often.
In one embodiment of this invention, pair bandwidth provider interchanger networked with data center progress PING surveys
Examination, including one or more interconnection ports of the bandwidth provider interchanger are tested.
In one embodiment of this invention, the fringe node for server, personal computer, tablet computer, smart mobile phone,
One of interchanger and router.
In one embodiment of this invention, the result that test is drawn is carried out to each controllable node includes following part or complete
Portion:Delay, packet loss, mdev variance yields and downloading rate value.
In one embodiment of this invention, the result that test is drawn is carried out to each controllable node includes delay and packet loss.
In one embodiment of this invention, the above method further includes obtains uncontrollable node as follows:Obtain region
Public IP storehouses;The public IP storehouses are sorted out;Common objective IP is tested in the test machine of diverse geographic location;Pass through ratio
To the test result of different test machine, stable public jump is found out;Public jump be referred to one belong to the IP geographical locations,
Operator, region stabilization IP storehouses in.
The present invention is due to using above technical scheme, being allowed to compared with prior art, have following remarkable advantage:The present invention
In addition to monitoring fringe node, also newly-increased monitoring data center intranet environment quality, monitoring test bandwidth provider core are handed over
End change planes to the network quality situation of node switching generator terminal, layering monitoring in this way can in the situation for network quality decline occur
Excluded with being layered, most fast orientation problem point.Also, the present invention except through monitoring controllable node outside, also monitor not
Controllable node, reaches the region that monitoring has monitoring blind spot by analysis monitoring result.
Brief description of the drawings
The above content of the invention and following embodiment of the present invention can obtain more preferably when being read in conjunction with the figure
Understanding.It should be noted that example of the attached drawing only as claimed invention.In the accompanying drawings, identical reference numeral
Represent same or similar element.
Fig. 1 shows a kind of typical network topology framework of data center in CDN network.
Fig. 2 shows the exemplary topology framework of CDN network.
Fig. 3 shows the layering monitoring flow of one embodiment of the invention.
Fig. 4 shows the fringe node monitoring flow of one embodiment of the invention.
Fig. 5 shows the monitoring method flow of one embodiment of the invention.
Embodiment
Describe the detailed features and advantage of the present invention in detail in a specific embodiment below, its content is enough to make any
Skilled in the art realises that the technology contents of the present invention and implement according to this, and specification according to disclosed by this specification, weigh
Profit requires and attached drawing, skilled person readily understands that the relevant purpose of the present invention and advantage.
Fig. 1 shows a kind of typical network topology framework of data center.In the network topology framework of the data center 100
With core switch 101 (such as switch A).120 (the example of interchanger that the core switch 101 is provided with service provider
Such as ISP interchangers) interconnected by one or more (3 are shown in figure) link.Core switch 101 can connect one or more
Second line of a couplet interchanger 102 (such as switch b).Node server 103 (such as cache A, cache B) in CDN network
It can be interconnected directly or by second line of a couplet interchanger 102 and core switch 101.Meanwhile the framework further includes at least one monitoring equipment
104, for monitor network quality.
Fig. 2 shows the exemplary topology framework of CDN network.Refering to what is shown in Fig. 2, the periphery connection of data center 100 is
Many fringe nodes 140.Fringe node 140 is located at outside data center 100, can be by being connected to the net of data center
Network equipment.For example, fringe node can be server, personal computer, tablet computer, smart mobile phone, interchanger, route
Device etc..
Usually, the data interaction of network can occur the core switch 101 in data center, second line of a couplet interchanger 102,
Between interchanger 120 and fringe node 140 that node server 103, service provider provide, thus need to test these equipment
The quality of the link of composition.
But known test is all immediately seen test result, the delay tested such as device A to B and packet loss, under
The time spent in carrying speed and the poll pages.But according to test result can not decision problem which process appeared in.For this reason,
According to an embodiment of the invention, there is provided the method that zonal testing is carried out according to test environment.Specifically, by CDN network
The link of each equipment room further decomposes, including data center's inner link, data center are exchanged with what service provider provided
The link of link and data center between machine 120 to fringe node.It is right using the monitoring equipment 104 inside data center
Each layer link carries out quality test, can understand the quality of each layer link respectively.
Further, since in fringe node, controllable number of nodes is extremely limited, in order to improve the covering surface of test
And accuracy, the embodiment of the present invention introduce uncontrollable node.Wherein controllable node is can be by the monitoring of data center
Person is by network entry and disposes the fringe node of the HTTP website services of downloading rate test and page Poll tests, can not
Node is controlled as neither downloading rate test and the page can not be disposed by the supervisor of data center by network entry
The fringe node of the HTTP website services of Poll tests.
Fig. 3 shows the layering monitoring flow of one embodiment of the invention.Refering to what is shown in Fig. 3, three layers of flow point is monitored.
Step 301 obtains the monitored object of three supervisory layers, such as interchanger, node server and fringe node.In step 302
Deng pending data, once receiving data, i.e., each layer monitoring is according to circumstances carried out.
If receiving the first layer data, data center's internal control is carried out in step 303, this is first layer monitoring.Tool
Say body, select a monitoring equipment 104 to be monitored inside data center 100, monitored object may include interchanger gateway IP
Address and all node server IP address.Here, test mode can be ping testing, including test monitoring object prolongs
When, the packet loss of monitored object.After if step 304 obtains data, you can draw test result into final step 309.
For test result, it can compare with one group of first reference value, to determine tested node server or clothes
Whether the interchanger that business device is connected is normal.For example, one group of first reference value is defined, be delayed 2ms, packet loss 0%.Packet loss
As long as rate is more than 0% or is delayed meets any one condition more than 2ms, then it is assumed that there are network problem inside data center.
If receiving the second layer data, carry out data center end in step 305 and exchange generator terminal monitoring to bandwidth provider,
This is second layer monitoring.Specifically, a monitoring equipment 104 is selected to be monitored inside data center 100;Determine that bandwidth carries
The one or more of generator terminal is exchanged for business can monitor IP address (generally equipment interconnection IP address);Then this IP address is tested.
Here, test mode can be ping testing, include the packet loss of test monitoring object.After if step 306 obtains data,
Test result can be drawn into final step 309.
, can be compared with one second reference value, to determine whether the bandwidth provider interchanger is normal for test result.
For example, an accuracy threshold, packet loss 0% are defined.If test result is more than threshold value packet loss 0%, then it is assumed that monitoring
Object has network problem.
If receiving third layer data, the monitoring of data center's end-to-edge node is carried out in step 307, this is the 3rd
Layer monitoring.Specifically, inside data center 100 monitoring equipment 104 can be selected to be monitored, monitored object includes can
Control node and uncontrollable node.After if step 308 obtains data, you can draw test result into final step 309.
, can be compared with standard value, to determine whether monitored object is normal for test result.For example, one is defined
A standard value, same to region (such as a province or city, similarly hereinafter) is with the downloading rate 1M/s of operator, and be delayed 60ms.The two ginsengs
Numerical value is threshold value, as long as downloading rate is less than 1Mbps or delay meets any one condition more than 60ms, thinks monitored object
There is network quality decline in end.
As it was noted above, the monitored object (in this fringe node) of third layer monitoring includes controllable node and uncontrollable section
Point.Fig. 4 shows the fringe node monitoring flow of one embodiment of the invention.Refering to what is shown in Fig. 4, in step 401, controllable node is obtained
With the test IP address of uncontrollable node.Here, controllable node is generally other data beyond the data center of monitoring equipment place
Node server in center.This kind of server 103 has an obvious feature, is controllable, and can dispose downloading rate
Test, page poll tests and ping testing.Uncontrollable node, generally by one stable covering IP storehouse of acquisition
Obtain.This kind of uncontrollable node is due to that cannot control, so most basic ping testing can only be provided, i.e., it is enabled that download is provided
And page poll tests are also a kind of to be considered as unstable test source.
After step 402 waits until data, i.e., 403 can be entered step, starts to test controllable and uncontrollable node.Controllable node
Can carry out ping testing, downloading rate test and page poll tests, tester can select one of them according to demand
Or all.For example, by downloading the test datas such as file record downloading rate.Uncontrollable node can be obtained by ping testing
Go out the test datas such as delay, packet loss.After if step 404 obtains data, you can draw test knot into final step 405
By.
For example, an accuracy threshold is defined, the downloading rate 1M/s with region with operator, be delayed 60ms, this two
A parameter value is threshold value, as long as downloading rate is less than 1Mbps or delay meets any one condition more than 60ms, thinks to monitor
There is network quality decline in residing data center.
In one embodiment of this invention, the acquisition methods of uncontrollable object approximately as:Obtain the public ip in big region
Storehouse;The IP storehouses of acquisition are sorted out according to certain refinement standard;Common mesh is tested in the test machine of diverse geographic location
Mark IP;By comparing different test machine test results, stable public jump is found out;Public jump is referred to one and new belongs to this
Ip geographical locations, operator, region stabilization ip storehouses in.
Fig. 5 shows the monitoring method flow of one embodiment of the invention.Refering to what is shown in Fig. 5, conclusion is described above, the present invention one
The monitoring method of embodiment comprises the following steps:
Step 501, the monitoring equipment inside data center carries out ping testing to the server of the data center;
Step 502, the bandwidth provider interchanger networked from the monitoring equipment pair inside data center with data center carries out
Ping testing;
Step 503, fringe node is tested from the monitoring equipment inside the data center, if wherein fringe node is
Controllable node, then carry out at least part of ping testing, downloading rate test and page Poll tests, if fringe node is not
Controllable node, then carry out ping testing.
The example of actual test is enumerated below.
First, the monitoring of data center switchboard internal soundness
Known monitoring mode can not monitor interchanger internal network quality, such as server load height, net card failure, cable
Internal network quality caused by loosening.
Such as Fig. 2, if real time monitoring arrives server (such as 103) all under notebook data center 100 on monitoring equipment 104
Delay, packet drop, it is possible to know the Network status inside interchanger.Can be by defining a standard delay value and one
Standard packet loss come weigh inside network condition.
Such as the delay of monitoring equipment 104 to cache A are 0.58ms, packet loss 0%;Monitoring equipment 104 is slow to high speed
The delay for depositing B is 3.75ms, packet loss 1%.
It is 2ms to define Intranet standard delay value, and standard packet loss is 0%.Can be with by the delay for testing Intranet machine
There is situations below:
1. other servers are arrived in monitoring equipment test, standard delay packet loss is occurred more than by three monitoring cycle portions machines
Situation, then the server is it is determined that be problematic machine, can allow artificial to go investigation be assorted by way of alarm
Caused by reason.
2. other servers are arrived in monitoring equipment test, standard delay is occurred more than by three monitoring cycle Servers-alls
The situation of packet loss, then can be determined that is the problem of monitoring machine, and same alarm, which out allows, artificially goes to investigate.
3. other servers are arrived in monitoring equipment test, showing the delay that is above standard by three monitoring cycle portions machines loses
The situation of bag, with reference to the machine position to go wrong as under same cascaded switches.It is in combination with the interchanger
It is no to occur running high or uneven situation, it is believed that further to investigate problem.
Two, data centers core switch are monitored to link-quality between bandwidth provider interchanger
As shown in Figure 1, the interchanger 120 of data center's core switch 101 and bandwidth provider interconnects, if wherein one
Bar link fails are as circuit loosening, the interruption of single-link load too high, single-link etc. will all cause network quality to be deteriorated
Consequence.If in the case of the intranet environment no problem within oneself interchanger is ensured, by going test core from monitoring equipment 104
Every upper level link situation of interchanger 101 can find failure caused by interconnection problem in time.
For example, the interchanger 120 that data center's core switch 101 and operator provide has three interconnection lines, right
The Internet address at end is respectively:192.168.1.1、192.168.1.2、192.168.1.3.
1. monitoring machine to test to interconnection port A, the feelings of standard delay packet loss are occurred more than by three monitoring cycles
Condition, then judge that interconnection port A is problematic, can allow artificial go caused by what reason investigation be by way of alarm.
Whether 2. being same as above, it is problematic equally can to test out other interconnection ports with such method.
As shown in Figure 1, in the place IP storehouses of a data center, there is N (N is positive integer) a IP.Monitoring system at present
The method of monitoring be test this IP places the downloading rate value of data center judge the network quality of the data center
Situation, if one's respective area with operator data center seldom or there is no when, the monitoring of the data center will deposit
In problem.
Quality-monitoring of the three, data centers to one's respective area IP storehouses
If the target of monitoring is not limited solely to one's respective area CDN data centers, but is expanded to the IP in one's respective area IP storehouses
Address, it is possible to more accurately monitor notebook data center comprehensively to live network quality.Specific method is as follows:
1. testing the IP address in the IP storehouses at the place of notebook data center 100 from the monitoring equipment 104 of data center 100, test can
Including ping testing and downloading rate test, page poll tests etc..
2. carry out ping testing, downloading rate test, page poll tests, no if IP address oneself data center
It then can only just carry out ping testing.
3. test need to draw the result is that packet loss, delay and mdev variance yields, if downloading rate value
Need to obtain.
4 values drawn and the standard value of definition are compared to weigh the network condition of this point.
Four, fringe nodes expand monitoring
There is one to cover certain domain test IP storehouses at present, IP address quantity is N number of, and only one of which IP address is IPA, this
IP is a server of data center, and the test being inclined to this when is exactly:
1. downloading rate is tested
[root@11011951~] #wget http://IPA:80/lvs.lxdns.net/test.rar-O/dev/
null
--2014-07-1110:04:20--http://IPA:80/lvs.lxdns.net/test.rar
Connecting to IPA:80...connected.
HTTP request sent,awaiting response...200OK
Length:2097152(2.0M)[application/octet-stream]
Saving to:`/dev/null'
100% [===============================
===========>]2,097,1524.20M/s in 0.5s
2. page poll is tested
[root@11011951~] #pollurl http://IPA:80/testcdn.htm;head-n
10pollUrl.temp
0
HTTP/1.0200OK
Content-Type:text/html
ETag:"2044265364"
Accept-Ranges:bytes
Last-Modified:Fri,06May 201107:59:23GMT
Content-Length:8
Connection:close
Date:Fri,11Jul 201402:04:27GMT
Server:WS CDN Server
These tests be all good at intuitively reaction test end to the tested access quality condition held how.
But if IPB is not the server of oneself, and simply one can be with the IP of ping testing.At this time just with delay
Go to weigh network quality with packet loss
3.PING is tested
[root@11011951~] #ping-fc 100IPB
PING IPB(IPB)56(84)bytes of data.
---111.206.217.15ping statistics---
100packets transmitted, 98received, 2%packet loss, time 1338ms
Rtt min/avg/max/mdev=22.706/29.328/41.211/3.073ms, pipe 4, ipg/
ewma13.517/28.441ms
A threshold value is drawn by a large amount of tests, can be with if delay is within 60ms or packet loss is within 1%
Meet requirement.It can thus make up because not being controllable server and test downloading rate or page cannot be removed
The shortcomings that face poll is tested, can equally go to obtain the network quality of overlay area, if an area of coverage by ping testing
Controllable node and the IP address of uncontrollable node are contained in the test IP storehouses in domain, can combine two kinds of test results and go to judge the overall situation
Network quality, accomplishes at utmost accurately to judge network quality.
In conclusion the embodiment of the present invention in addition to monitoring fringe node, also increases monitoring data center Intranet ring newly
Border quality, the network quality situation at monitoring test IDC bandwidth provider core switch ends to node switching generator terminal, in this way
Layering monitoring can be layered exclusion in the situation for network quality decline occur, most fast orientation problem point.Also, the present invention's
Embodiment is monitored outside speed except through downloading the file of controllable node, then is added monitoring and do not disposed monitored object region
IP storehouses, reach the region that monitoring has monitoring blind spot by analysis monitoring result.Pass through above method, the embodiment of the present invention
It is final to greatly improve monitoring covering surface, realize monitoring refinement layering, fast positioning trouble point.The method of the above embodiment of the present invention
It is adapted to use between the internal data center of CDN network, with the accuracy availability of ensuring method.
Although the present invention is described with reference to current specific embodiment, those of ordinary skill in the art
It should be appreciated that the embodiment of the above is intended merely to the explanation present invention, can also make in the case of without departing from spirit of the invention
Go out various equivalent change or replacement, therefore, as long as change, change in the spirit of the present invention to above-described embodiment
Type will all fall in the range of following claims.
Claims (7)
1. a kind of content distributing network quality control method, which includes data center, bandwidth provider exchanges
Machine and fringe node, this method comprise the following steps:
Monitoring equipment inside the data center carries out PING (Packet Internet to the node server of the data center
Grope) test;
The bandwidth provider interchanger networked from the monitoring equipment pair inside the data center with the data center carries out ping testing;
Fringe node is tested from the monitoring equipment inside the data center, if wherein the fringe node is controllable node,
At least part of ping testing, downloading rate test and page Poll tests is then carried out, if the fringe node is uncontrollable section
Point, then carry out ping testing;
Wherein controllable node for can by the supervisor of data center by network entry and dispose downloading rate test and page
The fringe node of the HTTP website services of face Poll tests, uncontrollable node are that can not pass through net by the supervisor of data center
Network logs in or cannot dispose the fringe node of the HTTP website services of downloading rate test and page Poll tests;
Uncontrollable node is wherein obtained in the following way:Obtain the public IP storehouses in region;The public IP storehouses are sorted out;Not
With test common objective IP in the test machine in geographical location;By comparing the test result of different test machine, stabilization is found out
Public jump;Public jump is referred to a stabilization for belonging to one or more of IP geographical locations, operator and the region
In IP storehouses.
2. content distributing network quality control method as claimed in claim 1, it is characterised in that
After the node server progress ping testing of the data center, the delay obtained and packet loss and one will be tested by further including
Just whether the first reference value of group compares, to determine interchanger that tested node server or node server connected
Often.
3. content distributing network quality control method as claimed in claim 1, it is characterised in that
Pair with the data center networking bandwidth provider interchanger carry out ping testing after, further include will test obtain packet loss
Rate is compared with one second reference value, to determine whether the bandwidth provider interchanger is normal.
4. content distributing network quality control method as claimed in claim 1, it is characterised in that pair with the data center network
Bandwidth provider interchanger carry out ping testing, including to one or more interconnection ports of the bandwidth provider interchanger into
Row test.
5. content distributing network quality control method as claimed in claim 1, it is characterised in that the fringe node is service
One of device, personal computer, tablet computer, smart mobile phone, interchanger and router.
6. content distributing network quality control method as claimed in claim 1, it is characterised in that surveyed to each controllable node
Try the result that draws include it is following part or all of:Delay, packet loss, mdev variance yields and downloading rate value.
7. content distributing network quality control method as claimed in claim 1, it is characterised in that surveyed to each controllable node
The result for trying to draw includes delay and packet loss.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410567526.8A CN104270291B (en) | 2014-10-22 | 2014-10-22 | CDN network quality control method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410567526.8A CN104270291B (en) | 2014-10-22 | 2014-10-22 | CDN network quality control method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104270291A CN104270291A (en) | 2015-01-07 |
CN104270291B true CN104270291B (en) | 2018-05-01 |
Family
ID=52161785
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410567526.8A Active CN104270291B (en) | 2014-10-22 | 2014-10-22 | CDN network quality control method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104270291B (en) |
Families Citing this family (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105187228A (en) * | 2015-06-12 | 2015-12-23 | 中国通信建设集团设计院有限公司 | Network quality detection method and router |
CN106851342A (en) * | 2017-03-06 | 2017-06-13 | 南京优速网络科技有限公司 | A kind of design architecture of content arrangement distribution network |
CN107181644B (en) * | 2017-06-16 | 2019-09-27 | 北京小度信息科技有限公司 | Method and apparatus for monitoring the network performance of network application |
CN107707378B (en) | 2017-06-29 | 2018-11-13 | 贵州白山云科技有限公司 | A kind of CDN covering scheme generation method and device |
CN107493192B (en) * | 2017-08-08 | 2018-10-16 | 深圳市网心科技有限公司 | Basic network method for evaluating quality and device for the transmission of video CD N business |
CN109962790B (en) * | 2017-12-14 | 2020-05-29 | 北京金山云网络技术有限公司 | Network quality monitoring method and device, electronic equipment and storage medium |
CN108063696A (en) * | 2017-12-14 | 2018-05-22 | 北京云端智度科技有限公司 | A kind of meshed network method for evaluating quality |
CN109462493B (en) * | 2018-09-13 | 2021-12-28 | 国网浙江省电力有限公司丽水供电公司 | Local area network monitoring method based on PING |
CN109347668B (en) * | 2018-10-17 | 2020-11-06 | 网宿科技股份有限公司 | Training method and device for service quality assessment model |
CN109450683A (en) * | 2018-11-09 | 2019-03-08 | 珠海格力电器股份有限公司 | Fault detection method, device, application system and storage medium |
CN109547276B (en) * | 2019-01-31 | 2022-04-12 | 网宿科技股份有限公司 | Problem positioning method, terminal and storage medium |
CN109981377B (en) * | 2019-04-10 | 2022-03-01 | 北京搜狐新媒体信息技术有限公司 | Distributed data center link monitoring method and system |
CN110225087A (en) * | 2019-05-08 | 2019-09-10 | 平安科技(深圳)有限公司 | Cloud access method, device and storage medium based on global load balancing |
CN110430092B (en) * | 2019-07-15 | 2021-03-16 | 杭州复杂美科技有限公司 | Download detection method, service node screening method, device and storage medium |
CN110505115B (en) * | 2019-07-30 | 2021-07-13 | 网宿科技股份有限公司 | Method and device for monitoring switch running-up risk |
CN110944061A (en) * | 2019-12-06 | 2020-03-31 | 北京博睿宏远数据科技股份有限公司 | Content distribution network scheduling method and device, computer equipment and storage medium |
CN111585833B (en) * | 2020-04-09 | 2022-03-11 | 新浪网技术(中国)有限公司 | Method and device for detecting public network quality of CDN node and computer equipment |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1054529A3 (en) * | 1999-05-20 | 2003-01-08 | Lucent Technologies Inc. | Method and apparatus for associating network usage with particular users |
CN101340312A (en) * | 2008-08-12 | 2009-01-07 | 腾讯科技(深圳)有限公司 | Network quality monitoring method and system of internet data center |
CN103036733B (en) * | 2011-10-09 | 2016-07-06 | 上海市南电信服务中心有限公司 | Unconventional network accesses monitoring system and the monitoring method of behavior |
CN102647299B (en) * | 2012-04-24 | 2014-10-15 | 网宿科技股份有限公司 | Hierarchical alarm analysis method and hierarchical alarm analysis system based on content delivery network |
CN103974139B (en) * | 2014-05-20 | 2017-02-15 | 北京速能数码网络技术有限公司 | Method and device for detecting network faults of terminal and system |
-
2014
- 2014-10-22 CN CN201410567526.8A patent/CN104270291B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN104270291A (en) | 2015-01-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104270291B (en) | CDN network quality control method | |
CN112087312B (en) | Method, device and equipment for providing edge service | |
US20200344199A1 (en) | Content node network address selection for content delivery | |
CN103795805B (en) | Distributed server load-balancing method based on SDN | |
US9729414B1 (en) | Monitoring service availability using distributed BGP routing feeds | |
KR101177203B1 (en) | Mapless global traffic load balancing via anycast | |
CN105340244B (en) | The method and apparatus of dynamic content dispensing network selection based on the context from transient state criterion | |
Wichtlhuber et al. | An SDN-based CDN/ISP collaboration architecture for managing high-volume flows | |
CN104272708B (en) | It is distributed with the stateless first order grouping to server farm and is distributed to the secondary data packets of the stateful second level grouping distribution of some server in group | |
CN102064992B (en) | Relay node, and relay node distributed network and networking method thereof | |
US9052955B2 (en) | System and method for seamless application hosting and migration in a network environment | |
CN107925681A (en) | The system and method that the load balance in network packet core system is defined for distributed software | |
CN104836732B (en) | The automatic selecting method and system of network connection | |
US20150081762A1 (en) | Systems and methods for distributing network traffic between servers based on elements in client packets | |
US9363313B2 (en) | Reducing virtual IP-address (VIP) failure detection time | |
CN103475654B (en) | Network path optimization method, equipment and network system | |
CN109962790A (en) | A kind of network quality monitoring method, device, electronic equipment and storage medium | |
CN111182022B (en) | Data transmission method and device, storage medium and electronic device | |
CN101217508A (en) | A network agent system and the corresponding realizing methods based on instant communication platform | |
US11457073B1 (en) | Supernode graceful shutdown in a proxy infrastructure | |
CN106302168A (en) | A kind of ISP route selecting method, device and gateway | |
VanderSloot et al. | Running refraction networking for real | |
Simoens et al. | Service-centric networking for distributed heterogeneous clouds | |
Lai et al. | Using SDN and NFV to enhance request rerouting in ISP-CDN collaborations | |
CN104270312B (en) | Support relay route distribution system and method that flow optimization and application are perceived |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right |
Effective date of registration: 20200706 Address after: 100083 Beijing Haidian District 39 Xueyuan Road 6 real estate building. Patentee after: Beijing Mesh Technology Co.,Ltd. Address before: 200030 Shanghai city Xuhui District Xietu Road No. 2899 A Kuangchi Cultural Square Building 5 floor Patentee before: WANGSU SCIENCE & TECHNOLOGY Co.,Ltd. |
|
TR01 | Transfer of patent right |