CN103248523A - Network route collocation method used for calculating array operating system on large scale - Google Patents
Network route collocation method used for calculating array operating system on large scale Download PDFInfo
- Publication number
- CN103248523A CN103248523A CN201310163986XA CN201310163986A CN103248523A CN 103248523 A CN103248523 A CN 103248523A CN 201310163986X A CN201310163986X A CN 201310163986XA CN 201310163986 A CN201310163986 A CN 201310163986A CN 103248523 A CN103248523 A CN 103248523A
- Authority
- CN
- China
- Prior art keywords
- node
- routing table
- compressed
- network
- file
- 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.)
- Granted
Links
Images
Landscapes
- Information Transfer Between Computers (AREA)
- Computer And Data Communications (AREA)
Abstract
The invention discloses a network route collocation method used for calculating an array operating system on a large scale. The method comprises the following implementation steps: 1) preparing an operating system inner core and an operating system image at serving nodes, naming the routing list documents of calculating nodes corresponding to network IDs of the calculating nodes by using monotone increasing continuous numbers, carrying out double-compression according to the preset fixed interval by adopting double compressed format, and gaining a plurality of segmentation compressed documents to be stored in the operating system image, 2) reading the network IDs of the nodes after the calculating nodes enter into the communication equipment configuration stage, acquiring the segmentation compressed documents corresponding to the nodes in the operating system image, uncompressing the segmentation compressed documents, acquiring the routing list documents of the nodes, and finishing the network route collocation according to the routing list documents. The method has the advantages that the occupied space of the routing list documents is small; the search of the needed routing list documents by the calculating nodes is fast; the network equipment configuration is fast; the system start time is short; and the maintainability and the usability are good.
Description
Technical field
The present invention relates to system-computed knot-net routing management, the configuration start-up technique field of high-performance computer system, be specifically related to a kind of network method for configuring route for large-scale calculations array processing system.
Background technology
High-performance computer system is made up of computing array, communication array, array of services usually, must be by high performance communication network connection, competence exertion effect between these arrays.At present, part high performance communication network is taked the source routing mode, through facts have proved that this mode has very high cost performance.In this case, require to be equipped with independent routing table for each terminal (as calculating node, service station) that participates in communicating by letter.Routing table is a text, the routing iinformation that has comprised certain network terminal, the routing table file is stored in the file system on the disk of node usually, but computing array quantity is huge, usually about 10,000, therefore, consider cost, process structure, the stability of a system, calculate node and be not equipped with hard disk.
Calculating node all is to read kernel and system image by network from service station, starts operation then, finishes a series of initialization, configuration effort.Wherein, must use the high-speed communication hardware that correct routing table file could correct configuration node, thus with other node proper communications of system.Therefore, efficiently, the routing table of the no disc gage calculation of configuration correctly node is the committed step that high-performance computer system starts.Each routing table file name, content have nothing in common with each other, distinguished by a unique network physical sign (hereinafter to be referred as network ID), when calculating node participation network service, use network ID to identify the position of this node in network, therefore it also is corresponding one by one with network ID calculating node, thereby corresponding one by one with the routing table file.
But, problem is under this express network framework, calculate before the node startup, network can only be implemented simple original message communicating, can't transmit big text, and, in advance the routing table file is all sent and be stored in certain piece memory block of node, be inappropriate for hardware designs, because this will bring the logic of extra complexity for communication hardware, take the internal memory of valuable communication hardware, also require simultaneously the BIOS of calculating node and the driver of operating system are proposed functional requirement, certainly will increase the complexity of whole system like this, cost and prolongation debug time.Therefore, can only use simple message communicating, simple routing iinformation (only comprising network ID, service station to the route of destination node) be sent to target calculate on the node.
Like this, intactly transmit kernel and no disk image that routing table file task has just been given operating system, wherein relevant with network configuration necessary step has:
1, all routing table files that calculate node all is put in the no disk image;
When 2, calculating the node startup, according to simple routing iinformation, from service station read operation system kernel and no disk image;
3, calculate the node kernel and start kernel, from no disk image, read the routing table file of needs according to known network ID, finish the work of configure network devices.
But aforesaid operations is faced with following problem:
1, the occupation space of routing table file increases and increases along with calculating the node scale, facts have proved that this increase is ultra linear, when to 8000 left and right sides, all routing table files surpass 1.6GB, have about 192MB after using instrument compression such as gzip, and operating system does not have the disk image maximum and has only 512MB yet, and will deposit necessary system file, usually can only remain about 100MB, therefore, when the calculating node is larger, can not fully guarantee intactly to deposit the routing table file;
2, namely allow to deposit down the routing table file by compress mode, when the calculating node is sought routing table file that needs from compressed file, decompression time is longer relatively, considerably beyond the temporal summation of other parts in the node start-up course, this with light, simplify that to be that the calculating node of purpose starts strategy conflicting, and had a strong impact on the availability of whole system.
Summary of the invention
Technical problem solved by the invention provide a kind of routing table file take up room little, calculate that node searches that required routing table file is fast, configure network devices quick, system start-up time is short, the property tieed up and the good network method for configuring route that is used for large-scale calculations array processing system of availability.
In order to solve the problems of the technologies described above, the technical solution used in the present invention is:
A kind of network method for configuring route for large-scale calculations array processing system, implementation step is as follows:
1) service station beamhouse operation system kernel and operation system image, the corresponding network ID that calculates node of routing table file that simultaneously each is calculated node uses the monotonically increasing continuous number to name, adopting two compressed formats to carry out dual compression according to default fixed step size all routing table files obtains a plurality of cutting compressed files, described a plurality of cutting compressed files is stored in the described operation system image;
2) calculate node and obtain operating system nucleus and operation system image to service station, kernel and enter communication device configured after the stage starts the operating system, read the network ID of this node, according to described fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image, with the routing table file that described cutting compressed file decompresses and obtains this node, finish the network routing configuration of this node according to described routing table file.
As further improvement in the technical proposal of the present invention:
The detailed step of described step 1) is as follows:
1.1) service station beamhouse operation system kernel and operation system image; Prepare the routing table file that all calculate node, use the monotonically increasing continuous number to name each corresponding network ID that calculates node of routing table file that calculates node;
1.2) use earlier gz form, back to use the bz2 form to carry out all routing table files of dual compression, record the compressed file size of all routing table files after through twice compression;
1.3) use earlier the bz2 form, the back uses the gz form to carry out all routing table files of dual compression, records the compressed file size of all routing table files after through twice compression;
1.4) comparison step 1.2) and step 1.3) two kinds of compressed file sizes that compressed format obtains in proper order, select the less compressed format order of compressed file as selected compressed format order;
1.5) serve as at interval all routing table files to be carried out segmentation with default fixed step size, adopt described selected compressed format to carry out dual compression in proper order all compressed files in each segmentation and obtain a cutting compressed file, finally obtain a plurality of cutting compressed files of corresponding a plurality of segmentations respectively;
1.6) described a plurality of cutting compressed files are stored in the described operation system image.
Described step 2) detailed step is as follows:
2.1) calculate node and obtain operating system nucleus and operation system image to service station, start the operating system kernel and enter communication device configured after the stage redirect carry out next step;
2.2) calculate the network ID that node reads this node;
2.3) calculate node and according to described fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image;
2.4) calculate node according to the backward of described selected compressed format order, adopt decompress described cutting compressed file and obtain the routing table file of this node of the order of pipe method;
2.5) calculate node is finished this node according to described routing table file network routing configuration;
2.6) service station calculates nodes at all and complete successfully all cutting compressed files in the deletion action system image after the network routing configuration.
The present invention has following advantage:
1, prior art the very corn of a subject is that superlinearity increases the routing table document size along with the growth of node scale, after facts have proved compression such as the compression instrument that uses gzip, the routing table document size that obtains is compared with the maximum (512MB) of no disc system reflection, substantially be same rank, this has had a strong impact on the placement of the system file of other kinds necessity, thereby directly causes system can't start use; The present invention carries out all routing table files dual compression and obtains a plurality of cutting compressed files according to default fixed intervals, the two compressed formats of employing, when calculating node number progressively increases, it is very little that the routing table file adopts two compressed formats to carry out the size increase that dual compression obtains a plurality of cutting compressed files, can too much not take no disc system map space.Experiment shows, when calculating node scale is 8000 (scale that large scale system is commonly used), after adopting two compressed formats, the routing table file compression rate can reach 20:1, and single compressed format has only 8:1, when original size is big, when namely radix is big, compare with the maximum (512MB) of no disc system reflection, the effect that two compressed formats are brought is very tangible.
2, the present invention calculates the node startup and enters communication device configured after the stage, obtain the cutting compressed file of corresponding this node on the service station according to the network ID of this node, the routing table file that the cutting compressed file is decompressed and obtains this node, finish the network routing configuration according to the routing table file that obtains, therefore, calculate node at each, it only is cutting compressed file rather than all a routing table file that comprises part routing table file by the network transmission, and the transmitted data on network amount still less; And the cutting compressed file rather than all the routing table files that also only need one of decompress(ion) to comprise part routing table file when calculating the node decompress(ion) get final product, greatly shortened the time that twice decompress(ion) spends, shortened and calculated node reads the routing table file from compressed file time.Experiment shows, when calculating node scale is 8000 (scale that large scale system is commonly used), prior art reads the used time of file from whole compressed file be 40 seconds, and the present invention is from being that to read the used time the compressed file of step-length cutting only be 0.6 second with 128, therefore can raise speed 66 times, effect also is very tangible.
Description of drawings
Fig. 1 is for using the large-scale calculations array processing system topology schematic diagram of the embodiment of the invention.
Fig. 2 is the holistic approach schematic flow sheet of the embodiment of the invention.
Fig. 3 is the method detailed schematic flow sheet of step 1) in the embodiment of the invention.
Fig. 4 is step 2 in the embodiment of the invention) the method detailed schematic flow sheet.
Embodiment
As shown in Figure 1, present embodiment is example with Milky Way high-performance computer system as typical large-scale calculations array processing system, comprise computing array and array of services in the Milky Way high-performance computer system, computing array links to each other by high-speed communicating network with array of services, computing array comprises a plurality of calculating nodes, array of services comprises a plurality of service stations, service station provides operating system nucleus and operation system image for calculating node, wherein comprise all routing table files in the operation system image, calculate node and then from server, read kernel and operation system image respectively, from operation system image, obtain the routing table file then.Large-scale calculations array processing system independently calculates node by several thousand usually and constitutes huge computing array, the calculating node of computing array must connect by complicated high speed internet network, can communicate by letter mutually, thereby the common large-scale concurrent program of cooperating operation, routing table file of each calculating node configuration could calculate node with other and communicate, but calculating node is the system of no dish, therefore need not have configuration routing table file in the disk image in operating system.
As shown in Figure 2, present embodiment is as follows for the step of the network method for configuring route of large-scale calculations array processing system:
1) service station beamhouse operation system kernel and operation system image, the corresponding network ID that calculates node of routing table file that simultaneously each is calculated node uses the monotonically increasing continuous number to name, adopting two compressed formats to carry out dual compression according to default fixed step size all routing table files obtains a plurality of cutting compressed files, a plurality of cutting compressed files is stored in the operation system image;
2) calculate node and obtain operating system nucleus and operation system image to service station, kernel and enter communication device configured after the stage starts the operating system, read the network ID of this node, according to fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image, with the routing table file that the cutting compressed file decompresses and obtains this node, finish the network routing configuration of this node according to the routing table file.
Present embodiment carries out two compressed format compressions of step 1) to the route list file, step 2) the cutting compressed file based on the segmentation cutting obtains the routing table file fast, adopt two compressed format compression routing table files by step 1), greatly slow down because calculating the superlinearity growth that the node scale increases the routing table file overall size that causes, by step 2) obtain the routing table file fast based on the cutting compressed file of segmentation cutting, greatly shorten and calculate the time that node is searched required routing table file, thereby make operating system configure network devices rapidly, greatly shorten system start-up time, improved the system-wide property tieed up and availability.
As shown in Figure 3, the detailed step of step 1) is as follows:
1.1) service station beamhouse operation system kernel and operation system image; Prepare the routing table file that all calculate node, use the monotonically increasing continuous number to name each corresponding network ID that calculates node of routing table file that calculates node.In the step 1) of present embodiment, service station is placed into the routing table file that each calculates node under the independent catalogue, and use monotonically increasing continuous number to comprise corresponding network ID in the middle of naming to guarantee in these filenames each corresponding network ID that calculates node of routing table file that calculates node, its concrete mask form is " rtb-n ", (n=0,1 ..., max calculation node number).Wherein, character string rtb is that routing table file (route table) is called for short, n is network ID, n represents with 16 systems in the present embodiment, as 0x0,0xf, 0x123 etc., for example be expressed as 0xc67 for network id for the 3175(16 system) the calculating node for, its routing table file format is " rtb-0xc67 ".
1.2) use earlier gz form, back to use the bz2 form to carry out all routing table files of dual compression, record the compressed file size of all routing table files after through twice compression.In the present embodiment, use earlier gz form, back use compress order that the bz2 form carries out all routing table files of dual compression for " tar czf – rtb-* | bzip2 〉/tmp/rtb.tgz.bz2 ".
1.3) use earlier the bz2 form, the back uses the gz form to carry out all routing table files of dual compression, records the compressed file size of all routing table files after through twice compression; In the present embodiment, use earlier the bz2 form, the back use compress order that the gz form carries out all routing table files of dual compression for " bzip2 – rtb-* | tar czf 〉/tmp/rtb.bz2.gz ".
1.4) comparison step 1.2) and step 1.3) two kinds of compressed file sizes that compressed format obtains in proper order, select the less compressed format order of compressed file as selected compressed format order; Because different its compression ratios of compressed format order also can be distinct, therefore present embodiment compresses in proper order by selected compressed format, can preferentially select the littler selected compressed format of compressed format to realize compression to the route list file in proper order, thereby can realize routing table file huge in the large-scale calculations array normally directly is stored in the no disc system reflection, only take less space, the system that guaranteed normally starts operation.
1.5) serve as at interval all routing table files to be carried out segmentation with default fixed step size, adopting selected compressed format to carry out dual compression in proper order all compressed files in each segmentation obtains a cutting compressed file, finally obtains a plurality of cutting compressed files of corresponding a plurality of segmentations respectively; In the present embodiment, fixed step size is chosen as 128, so the 0th～127 compressed 1 cutting compressed file " rtb-1.tgz.bz2 " that obtains of routing table file that calculates node, the 128th～255 compressed 1 the cutting compressed file " rtb-2.tgz.bz2 " that obtains of routing table file that calculates node, and the like.
1.6) store a plurality of cutting compressed files in the operation system image (under/etc/route catalogue).
After calculating the node startup, need from the routing table file of compression, find the routing table file of its network ID correspondence, because the routing table file of present embodiment is taked two compressed formats, when file is big, need implement decompress(ion) twice to whole file, required time is considerably beyond the start-up time of other parts.For only to improve the decompress(ion) speed of two compressed format routing table files, make and both can guarantee that the routing table file took less space, can obtain decompress(ion) speed faster again, network ID is continuous digital form in the present embodiment, thereby the routing table file also is to name with the continuous numeral of monotonically increasing, and only need obtain a routing table file at every turn, therefore present embodiment is a plurality of cutting compressed files with the routing table file of compression according to the fixed intervals cutting, calculate node and determine corresponding cutting compressed file enforcement decompress(ion) according to its network ID, required time reduces superlinearity ground.
As shown in Figure 4, detailed step step 2) is as follows:
2.1) calculate node and obtain operating system nucleus and operation system image to service station, start the operating system kernel and enter communication device configured after the stage redirect carry out next step.
2.2) calculate the network ID that node reads this node.
2.3) calculate node and according to fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image; Present embodiment is by the two compressed format files of segmentation cutting, and ID Network Based selects one of them cutting compressed file, so that cutting compressed file of follow-up decompress(ion) obtains the routing table file corresponding with network ID, greatly reduce the start-up time of calculating node, improved the system-wide property tieed up, availability.
Be 0xc67(namely 3175 with the network ID) the calculating node be example, calculate node and carry out modulo operation according to the network ID (3175) of 128 pairs of these nodes of fixed step size and obtain 24, namely the rtb file of node 0xc67 (routing table file) is in cutting compressed file rtb-24.tgz.bz2; Then this calculating node is according to the 24 cutting compressed files " rtb-24.tgz.bz2 " that find corresponding this node in operation system image.
2.4) calculate node according to the backward of selected compressed format order, adopt the order decompression cutting compressed file of pipe method and obtain the routing table file of this node; Present embodiment adopts the order decompression cutting compressed file of pipe method not take the operation system image space, can reduce decompression cutting compressed file taking the operation system image space.
Behind the cutting compressed file " rtb-24.tgz.bz2 " that finds corresponding this node, decompressing obtains the rtb-24.tgz.bz2 decompress(ion) routing table file " rtb-0xc67 " of this node to calculate node fill order " bzcat rtb-24.tgz.bz2 | tar zxf – rtb-0xc67 – C/etc/pdp.rtb ", and is put under designated directory/etc/ and gets final product.
2.5) calculate node is finished this node according to the routing table file network routing configuration.
2.6) service station calculates nodes at all and complete successfully after the network routing configuration all cutting compressed files of (under/etc/route catalogue) in the deletion action system image, thereby discharge more multisystem space.
The above only is preferred implementation of the present invention, and protection scope of the present invention also not only is confined to above-described embodiment, and all technical schemes that belongs under the thinking of the present invention all belong to protection scope of the present invention.Should be pointed out that for those skilled in the art in the some improvements and modifications that do not break away under the principle of the invention prerequisite, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (3)
1. network method for configuring route that is used for large-scale calculations array processing system is characterized in that implementation step is as follows:
1) service station beamhouse operation system kernel and operation system image, the corresponding network ID that calculates node of routing table file that simultaneously each is calculated node uses the monotonically increasing continuous number to name, adopting two compressed formats to carry out dual compression according to default fixed step size all routing table files obtains a plurality of cutting compressed files, described a plurality of cutting compressed files is stored in the described operation system image;
2) calculate node and obtain operating system nucleus and operation system image to service station, kernel and enter communication device configured after the stage starts the operating system, read the network ID of this node, according to described fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image, with the routing table file that described cutting compressed file decompresses and obtains this node, finish the network routing configuration of this node according to described routing table file.
2. the network method for configuring route for large-scale calculations array processing system according to claim 1 is characterized in that the detailed step of described step 1) is as follows:
1.1) service station beamhouse operation system kernel and operation system image; Prepare the routing table file that all calculate node, use the monotonically increasing continuous number to name each corresponding network ID that calculates node of routing table file that calculates node;
1.2) use earlier gz form, back to use the bz2 form to carry out all routing table files of dual compression, record the compressed file size of all routing table files after through twice compression;
1.3) use earlier the bz2 form, the back uses the gz form to carry out all routing table files of dual compression, records the compressed file size of all routing table files after through twice compression;
1.4) comparison step 1.2) and step 1.3) two kinds of compressed file sizes that compressed format obtains in proper order, select the less compressed format order of compressed file as selected compressed format order;
1.5) serve as at interval all routing table files to be carried out segmentation with default fixed step size, adopt described selected compressed format to carry out dual compression in proper order all compressed files in each segmentation and obtain a cutting compressed file, finally obtain a plurality of cutting compressed files of corresponding a plurality of segmentations respectively;
1.6) described a plurality of cutting compressed files are stored in the described operation system image.
3. the network method for configuring route for large-scale calculations array processing system according to claim 2 is characterized in that described step 2) detailed step as follows:
2.1) calculate node and obtain operating system nucleus and operation system image to service station, start the operating system kernel and enter communication device configured after the stage redirect carry out next step;
2.2) calculate the network ID that node reads this node;
2.3) calculate node and according to described fixed step size the network ID of this node is carried out the cutting compressed file that modulo operation obtains corresponding this node in the operation system image;
2.4) calculate node according to the backward of described selected compressed format order, adopt decompress described cutting compressed file and obtain the routing table file of this node of the order of pipe method;
2.5) calculate node is finished this node according to described routing table file network routing configuration;
2.6) service station calculates nodes at all and complete successfully all cutting compressed files in the deletion action system image after the network routing configuration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310163986.XA CN103248523B (en) | 2013-05-07 | 2013-05-07 | Network route collocation method used for calculating array operating system on large scale |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310163986.XA CN103248523B (en) | 2013-05-07 | 2013-05-07 | Network route collocation method used for calculating array operating system on large scale |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103248523A true CN103248523A (en) | 2013-08-14 |
| CN103248523B CN103248523B (en) | 2014-04-16 |
Family
ID=48927755
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310163986.XA Active CN103248523B (en) | 2013-05-07 | 2013-05-07 | Network route collocation method used for calculating array operating system on large scale |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103248523B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101088084A (en) * | 2003-12-29 | 2007-12-12 | 文丘里无线公司 | Reusable compressed objects |
| CN102447604A (en) * | 2010-09-30 | 2012-05-09 | 迈普通信技术股份有限公司 | Storage method for routing table information and route equipment |
| CN102708574A (en) * | 2011-02-25 | 2012-10-03 | 奥多比公司 | Compression of image data |
| CN102957915A (en) * | 2012-11-15 | 2013-03-06 | 西安理工大学 | Double JPEG (Joint Photographic Experts Group) compressed image-targeted tempertamper detection and tempertamper locating method |
-
2013
- 2013-05-07 CN CN201310163986.XA patent/CN103248523B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101088084A (en) * | 2003-12-29 | 2007-12-12 | 文丘里无线公司 | Reusable compressed objects |
| CN102447604A (en) * | 2010-09-30 | 2012-05-09 | 迈普通信技术股份有限公司 | Storage method for routing table information and route equipment |
| CN102708574A (en) * | 2011-02-25 | 2012-10-03 | 奥多比公司 | Compression of image data |
| CN102957915A (en) * | 2012-11-15 | 2013-03-06 | 西安理工大学 | Double JPEG (Joint Photographic Experts Group) compressed image-targeted tempertamper detection and tempertamper locating method |
Non-Patent Citations (2)
| Title |
|---|
| 周双英， 余建桥: "RWM & DEWS数据二次压缩算法研究", 《计算机工程》 * |
| 张丽霞，宋鸿陟: "多重压缩DNA序列数据", 《计算机应用》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103248523B (en) | 2014-04-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP2948875B1 (en) | Method and system for using a recursive event listener on a node in hierarchical data structure | |
| CN107977396B (en) | Method and device for updating data table of KeyValue database | |
| CN103078907B (en) | Upload, cloud backs up, search, recover method and the device of data | |
| US8838550B1 (en) | Readable text-based compression of resource identifiers | |
| CN105162855A (en) | Incremental data synchronization method and device | |
| CN111104069A (en) | Multi-region data processing method and device of distributed storage system and electronic equipment | |
| CN105740405B (en) | Method and device for storing data | |
| CN110727738B (en) | Global routing system based on data fragmentation, electronic equipment and storage medium | |
| CN102946410A (en) | Method and device for network synchronization | |
| CN109753424B (en) | AB test method and device | |
| CN110851525A (en) | Data fragmentation method, related equipment and computer storage medium | |
| CN110460668B (en) | File uploading method and device, computer equipment and storage medium | |
| CN103500129A (en) | Back-up object sending and back-up method, production end, backup-for-disaster-recovery end and system | |
| CN114398520A (en) | Data retrieval method, system, device, electronic equipment and storage medium | |
| CN104205730A (en) | Network element data access method, device and network management system | |
| WO2017206562A1 (en) | Data table processing method, device, and system | |
| CN103248523B (en) | Network route collocation method used for calculating array operating system on large scale | |
| CN112988879A (en) | Method, system, device, storage medium and processor for accessing database | |
| CN106919574B (en) | Method for processing remote synchronous file in real time | |
| JP2015064636A (en) | Information processing system, distribution processing method, and program | |
| JP5756379B2 (en) | Data transfer system, transfer source system, transfer destination system, and program | |
| Bonsma et al. | Extremal graphs having no matching cuts | |
| CN111061719B (en) | Data collection method, device, equipment and storage medium | |
| CN110569456B (en) | WEB end data offline caching method and device and electronic equipment | |
| CN114138786A (en) | Method, device, medium, product and equipment for duplicate removal of online transaction message |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |