CN105373449A - Distributed storage metadata recovery method, apparatus and system - Google Patents

Distributed storage metadata recovery method, apparatus and system Download PDF

Info

Publication number
CN105373449A
CN105373449A CN201510799090.XA CN201510799090A CN105373449A CN 105373449 A CN105373449 A CN 105373449A CN 201510799090 A CN201510799090 A CN 201510799090A CN 105373449 A CN105373449 A CN 105373449A
Authority
CN
China
Prior art keywords
node
metadata
message
lacks
mirror
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
Application number
CN201510799090.XA
Other languages
Chinese (zh)
Other versions
CN105373449B (en
Inventor
冯永刚
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huawei Technologies Co Ltd
Original Assignee
Huawei Technologies Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Huawei Technologies Co Ltd filed Critical Huawei Technologies Co Ltd
Priority to CN201510799090.XA priority Critical patent/CN105373449B/en
Publication of CN105373449A publication Critical patent/CN105373449A/en
Application granted granted Critical
Publication of CN105373449B publication Critical patent/CN105373449B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)

Abstract

Embodiments of the invention provide a distributed storage metadata recovery method, apparatus and system, which can timely recover metadata of fault nodes in a scene with node faults, so that subsequent mirror writing operation does not generate degradation writing logs any more. The method comprises the steps that a first node receives a first message sent by a second node, wherein the first message carries first metadata and an operation request of performing mirror writing operation on the first metadata; if the metadata of the first node is incomplete, the first node obtains missing metadata of the first node; the first node recovers the metadata of the first node according to the obtained missing metadata; and after the metadata of the first node is recovered completely, the first node performs the mirror writing operation on the first metadata and sends a second message to the second node, wherein the second message is used for indicating that the mirror writing operation of the first metadata is successful. The distributed storage metadata recovery method, apparatus and system are suitable for the technical field of data storage.

Description

The restorative procedure of distributed storage metadata, Apparatus and system
Technical field
The present invention relates to technical field of data storage, particularly relate to the restorative procedure of distributed storage metadata, Apparatus and system.
Background technology
In existing distributed storage architecture, very high to the reliability requirement of metadata.Metadata is usually kept at different nodes in the form of a mirror image and backs up, to ensure the reliability of data.Under the scene having node failure, the mirror image reliability of metadata can reduce, and node recovers normally, and the metadata it being write failure will be repaired correctly, to ensure the reliability of metadata timely.
How to repair the metadata on once malfunctioning node, common mode be degradation when writing record to repair the daily record of object, wait for background server obtain repair daily record after carry out the reparation of metadata, and then ensure the reliability of metadata.
But, before the metadata reparation correctly that background server was write malfunctioning node was once demoted, if the metadata again on this malfunctioning node of mirror-write, because the metadata in the metadata on this malfunctioning node and other node is inconsistent, therefore will cause continuing degradation to write, and generate new degradation and write daily record.When the mirror-write operating frequency of metadata is higher, a large amount of degradation of generation is write daily record, thus adds the task amount of the metadata that background server will recover.The daily record loss write if demoted or daily record also do not perform, and at this time have other nodes break down again, distributed storage metadata may be caused to lose, and then lose user file.
Therefore, under the scene having node failure, how to repair the metadata on malfunctioning node in time, make follow-up mirror-write operation can not produce degradation again and write daily record, become current problem demanding prompt solution.
Summary of the invention
The embodiment of the present invention provides restorative procedure, the Apparatus and system of distributed storage metadata, can repair the metadata on malfunctioning node under the scene having node failure in time, makes follow-up mirror-write operation can not produce degradation again and writes daily record.
For achieving the above object, the embodiment of the present invention provides following technical scheme:
First aspect, provide a kind of restorative procedure of distributed storage metadata, described method comprises:
First node receives the first message that Section Point sends, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks;
Described first node, according to the metadata of the disappearance got, is repaired the metadata on described first node;
By on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and sends the second message to described Section Point, and described second message is used to indicate described first metadata mirror-write successful operation.
In the first possible implementation of first aspect, in conjunction with first aspect, described first node obtains the metadata that described first node lacks, and comprising:
Described first node sends the 3rd message to described Section Point, and described 3rd message is for asking the metadata that described first node lacks;
Described first node receives the 4th message that described Section Point sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
In the implementation that first aspect the second is possible, in conjunction with first aspect, described first node obtains the metadata that described first node lacks, and comprising:
Described first node sends the 3rd message to meta data server MDS, and described 3rd message is for asking the metadata that described first node lacks;
Described first node receives the 4th message that described MDS sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
Second aspect, provide a kind of restorative procedure of distributed storage metadata, described method comprises:
Section Point sends the first message to first node, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described Section Point receives the second message that described first node sends, and described second message is for asking the metadata that described first node lacks;
Described Section Point sends the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node;
Described Section Point receives the 4th message that described first node sends, and described 4th message is used to indicate described first metadata mirror-write successful operation.
In the first possible implementation of second aspect, in conjunction with second aspect, before described Section Point sends the first message to first node, also comprise:
Described Section Point receives described first message that meta data server MDS sends;
Described Section Point, according to described first message, carries out mirror-write operation to described first metadata.
The third aspect, provide a kind of restorative procedure of distributed storage metadata, described method comprises:
Meta data server MDS sends the first message to Section Point, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described MDS receives the second message that first node sends, and described second message is for asking the metadata that described first node lacks;
Described MDS sends the 3rd message to described first node, and described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Fourth aspect, provides a kind of first node, and described first node comprises: receiving element, processing unit and transmitting element;
Described receiving element, for receiving the first message that Section Point sends, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described processing unit, if the metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks;
Described processing unit, also for the metadata according to the disappearance got, repairs the metadata on described first node;
Described processing unit, also for by described first node metadata repair complete after, mirror-write operation is carried out to described first metadata;
Described transmitting element, also for sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
In the first possible implementation of fourth aspect, in conjunction with fourth aspect, described processing unit specifically for:
Send the 3rd message by described transmitting element to described Section Point, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described Section Point transmission by described receiving element, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
In the implementation that fourth aspect the second is possible, in conjunction with fourth aspect, described processing unit specifically for:
Send the 3rd message by described transmitting element to meta data server MDS, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described MDS transmission by described receiving element, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
5th aspect, provides a kind of Section Point, and described Section Point comprises: receiving element and transmitting element;
Described transmitting element, for sending the first message to first node, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiving element, if the metadata on described first node is imperfect, receive the second message that described first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitting element, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node;
Described receiving element, also for receiving the 4th message that described first node sends, described 4th message is used to indicate described first metadata mirror-write successful operation.
In the 5th in the first possible implementation, in conjunction with the 5th aspect, described Section Point also comprises processing unit;
Described receiving element, also for before sending the first message at described transmitting element to first node, receives described first message that meta data server MDS sends;
Described processing unit, for according to described first message, carries out mirror-write operation to described first metadata.
6th aspect, provide a kind of meta data server MDS, described MDS comprises: receiving element transmitting element;
Described transmitting element, for sending the first message to Section Point, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiving element, if the metadata on described first node is imperfect, receive the second message that first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitting element, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
7th aspect, provides a kind of first node, and described first node comprises: receiver, processor, transmitter and bus; Wherein, described receiver, described processor and described transmitter are connected by described bus and complete mutual communication;
Described receiver, for receiving the first message that Section Point sends, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described processor, if the metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks;
Described processor, also for the metadata according to the disappearance got, repairs the metadata on described first node;
Described processor, also for by described first node metadata repair complete after, mirror-write operation is carried out to described first metadata;
Described transmitter, also for sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
In the 7th in the first possible implementation, in conjunction with the 7th aspect, described processor specifically for:
Send the 3rd message by described transmitter to described Section Point, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described Section Point transmission by described receiver, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
In the implementation that the second is possible in the 7th, in conjunction with the 7th aspect, described processor specifically for:
Send the 3rd message by described transmitter to meta data server MDS, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described MDS transmission by described receiver, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
Eighth aspect, provides a kind of Section Point, and described Section Point comprises: receiver, transmitter and bus; Wherein, described receiver and described transmitter are connected by described bus and complete mutual communication;
Described transmitter, for sending the first message to first node, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiver, if the metadata on described first node is imperfect, receive the second message that described first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitter, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node;
Described receiver, also for receiving the 4th message that described first node sends, described 4th message is used to indicate described first metadata mirror-write successful operation.
In the first possible implementation of eighth aspect, in conjunction with eighth aspect, described Section Point also comprises processor;
Described receiver, also for before sending the first message at described transmitter to first node, receives described first message that meta data server MDS sends;
Described processor, for according to described first message, carries out mirror-write operation to described first metadata.
9th aspect, provide a kind of meta data server MDS, described MDS comprises: receiver, transmitter and bus; Wherein, described receiver and described transmitter are connected by described bus and complete mutual communication;
Described transmitter, for sending the first message to Section Point, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiver, if the metadata on described first node is imperfect, receive the second message that first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitter, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Tenth aspect, provides a kind of distributed storage architecture, comprises the first node as described in any one of fourth aspect, the Section Point as described in the 5th any one of aspect and the meta data server MDS as described in the 6th aspect;
Or, comprise the first node as described in the 7th any one of aspect, the Section Point as described in any one of eighth aspect and the MDS as described in the 9th aspect.
Based on the restorative procedure of the distributed storage metadata that the embodiment of the present invention provides, Apparatus and system, in the embodiment of the present invention, after first node receives first message of operation requests of carrying the first metadata and the first metadata being carried out to mirror-write operation that Section Point sends, when metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks, and then according to the metadata of the disappearance got, the metadata on described first node is repaired, by on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and send successful second message of the described first metadata mirror-write of instruction to Section Point.That is, and unlike in prior art, when the metadata on first node is imperfect, need to wait for the reparation carrying out metadata after background server obtains reparation daily record, when the mirror-write operating frequency of metadata is higher, a large amount of degradation daily records may be generated.In the embodiment of the present invention, can metadata on first node imperfect time, in time the metadata on first node is repaired, makes this mirror-write successful operation, can not generate new degradation and write daily record, and then the follow-up more degradation that also can not generate writes daily record.Further, because the follow-up more degradation that can not generate writes daily record, therefore it also avoid in prior art when generate a large amount of degradations write daily record time, distributed storage metadata may be caused to lose, and then lose the problem of user file, add the reliability of metadata mirror image in distributed storage architecture.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
The distributed storage architecture schematic diagram that Fig. 1 provides for the embodiment of the present invention;
The storage means of distributed storing metadata when Fig. 2 is prior art interior joint fault;
The restorative procedure of distributed storing metadata when Fig. 3 is the fault recovery of prior art interior joint;
The mutual schematic diagram of restorative procedure of the distributed storage metadata that Fig. 4 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet one of the distributed storage metadata that Fig. 5 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet two of the distributed storage metadata that Fig. 6 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet three of the distributed storage metadata that Fig. 7 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet four of the distributed storage metadata that Fig. 8 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet five of the distributed storage metadata that Fig. 9 provides for the embodiment of the present invention;
The restorative procedure schematic flow sheet six of the distributed storage metadata that Figure 10 provides for the embodiment of the present invention;
The first node structural representation one that Figure 11 provides for the embodiment of the present invention;
The Section Point structural representation one that Figure 12 provides for the embodiment of the present invention;
The Section Point structural representation two that Figure 13 provides for the embodiment of the present invention;
The MDS structural representation one that Figure 14 provides for the embodiment of the present invention;
The first node structural representation two that Figure 15 provides for the embodiment of the present invention;
The Section Point structural representation three that Figure 16 provides for the embodiment of the present invention;
The Section Point structural representation four that Figure 17 provides for the embodiment of the present invention;
The MDS structural representation two that Figure 18 provides for the embodiment of the present invention.
Embodiment
In order to being described clearly of following each embodiment is succinct, first provide briefly introducing of related notion:
The first, metadata
Metadata, refers to the data for data of description and environment thereof, and such as, metadata may be used for size, the access time and modification time etc. of data of description.
The second, mirror image is (English: Mirroring)
Mirror image is a type of redundancy, and the data on a disk exist an identical copy and are mirror image on another disk.In the embodiment of the present invention, mirror-write operation is, and identical data is write in another disk.
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
It should be noted that, for the ease of the technical scheme of the clear description embodiment of the present invention, in an embodiment of the present invention, have employed the printed words such as " first ", " second " to distinguish the substantially identical identical entry of function and efficacy or similar item, it will be appreciated by those skilled in the art that the printed words such as " first ", " second " do not limit quantity and execution order.
The embodiment of the present invention is mainly used in distributed storage architecture as shown in Figure 1.As shown in Figure 1, in this distributed storage architecture, meta data server is (English: MetaDataServer, be called for short: multiple nodes MDS) in unified management node pool, MDS can select certain node in node pool to be management node, and the Data dissemination that this management node is responsible for MDS to issue is on other node in node pool.When this management node breaks down, the MDS another one node that can reselect in node pool is management node, namely usually said management node " drift ".
Wherein, the node in the embodiment of the present invention can be specifically some disks, and the present invention does not do concrete restriction to this.
It should be noted that, in the distributed storage architecture shown in Fig. 1, MDS is independent of the node deployment in node pool.Certainly, MDS can also be deployed on the some nodes in node pool, and the embodiment of the present invention does not do concrete restriction to this.
Based on the distributed storage architecture shown in Fig. 1, suppose that node 3 is for management node, node 1, node 2, node 4 and node 5 is metadata mirror nodes, if when then node 5 breaks down (node 5 is malfunctioning node), then node 3 receive MDS send request metadata 1 is carried out to the request message 1 of mirror-write operation after, node 1, node 2, node 3, node 4 and node 5, by the storage means of execution distributed storage metadata as shown in Figure 2, comprising:
S201, node 3 send message 1 to node 1, and described message 1 carries metadata 1, and carry out the operation requests of mirror-write operation to metadata 1.
S202, node 3 send message 1 to node 2, and described message 1 carries metadata 1, and carry out the operation requests of mirror-write operation to metadata 1.
S203, node 3 send message 1 to node 4, and described message 1 carries metadata 1, and carry out the operation requests of mirror-write operation to metadata 1.
S204, node 3 send message 1 to node 5, and described message 1 carries metadata 1, and carry out the operation requests of mirror-write operation to metadata 1.
The described message 1 that S205, node 1 receiving node 3 send.
The described message 1 that S206, node 2 receiving node 3 send.
The described message 1 that S207, node 4 receiving node 3 send.
S208, node 1 send message 2 to node 3, and described message 2 is used to indicate metadata 1 mirror-write successful operation.
S209, node 2 send message 2 to node 3, and described message 2 is used to indicate metadata 1 mirror-write successful operation.
S210, node 4 send message 2 to node 3, and described message 2 is used to indicate metadata 1 mirror-write successful operation.
The described message 2 that S211, node 3 receiving node 1 send.
The described message 2 that S212, node 3 receiving node 2 send.
The described message 2 that S213, node 3 receiving node 4 send.
S214, node 3 generate the daily record that node 5 writes failure.
Because node 5 breaks down, therefore node 5 cannot receive the message 1 that node 3 sends, and then also can not send message 2 to node 3, and then node 3 cannot receive the message 2 of node 3 transmission, thus generates the daily record that node 3 writes failure.
It should be noted that, in the storage means of above-mentioned distributed meta data, inevitable execution sequencing is not had between step S201-S204, successively can perform any one step in step S201-S204, such as, perform step S201-S204 in turn, also can perform step S202, S203, S201, S204 etc. in turn, can also perform step S201-S204, the embodiment of the present invention does not do concrete restriction to this simultaneously.Do not have inevitable execution sequencing between step S205-S207, wherein, step S205 performs after step S201, and step S206 performs after step S202, and step S207 performs after step S203.Do not have inevitable execution sequencing between step S208-S210, wherein, step S208 performs after step S205, and step S209 performs after step S206, and step S210 performs after step S207.Do not have inevitable execution sequencing between step S211-S213, wherein, step S211 performs after step S208, and step S212 performs after step S209, and step S213 performs after step S210.
In prior art, generate after node 5 writes the daily record of failure because of the failure problems of node 5 at node 3, when background work personnel will repair node failure in time, after node failure Resolving probiems, how to repair the metadata on once malfunctioning node, common mode is the daily record of recording when writing and will repair object of demoting, and waits for the reparation carrying out metadata after background server obtains reparation daily record, and then ensures the reliability of metadata.As shown in Figure 3, comprising:
S301, node 3 send message 3 to node 1, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S302, node 3 send message 3 to node 2, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S303, node 3 send message 3 to node 4, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S304, node 3 send message 3 to node 5, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
The described message 3 that S305, node 1 receiving node 3 send.
The described message 3 that S306, node 2 receiving node 3 send.
The described message 3 that S307, node 4 receiving node 3 send.
The described message 3 that S308, node 5 receiving node 3 send.
S309, node 1 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
S310, node 2 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
S311, node 4 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
If the metadata on S312 node 5 is imperfect, node 5 sends message 5 to node 3, and described message 5 is used to indicate metadata 2 mirror-write operation failure.
Owing to when there is node failure in node 5, mirror-write failure is carried out to metadata 1, therefore after node 5 receives the message 3 of node 3 transmission, by imperfect for the metadata object judged on this node, and then mirror-write operation cannot be carried out to metadata 2, therefore node 5 sends message 5 to node 3, and described message 5 is used to indicate metadata 2 mirror-write operation failure.
The described message 4 that S313, node 3 receiving node 1 send.
The described message 4 that S314, node 3 receiving node 2 send.
The described message 4 that S315, node 3 receiving node 4 send.
The described message 5 that S316, node 3 receiving node 4 send.
S317, node 3 generate the daily record that node 5 writes failure.
S318, node 3 obtain the metadata of node 5 disappearance from background server.
Now, on node 5, the metadata of disappearance comprises metadata 1 and metadata 2.
The metadata that S319, node 3 lack on node 5 sending node 5.
S320, node 5, according to the metadata of the disappearance got, are repaired the metadata on node 5.
It should be noted that, in the restorative procedure of above-mentioned distributed meta data, inevitable execution sequencing is not had between step S301-S304, successively can perform any one step in step S301-S304, such as, perform step S301-S304 in turn, also can perform step S302, S303, S301, S304 etc. in turn, can also perform step S301-S304, the embodiment of the present invention does not do concrete restriction to this simultaneously.Do not have inevitable execution sequencing between step S305-S308, wherein, step S305 performs after step S301, and step S306 performs after step S302, and step S307 performs after step S303, and step S308 performs after step S304.Do not have inevitable execution sequencing between step S309-S312, wherein, step S309 performs after step S305, and step S310 performs after step S306, and step S311 performs after step S307, and step S312 performs after step S308.Do not have inevitable execution sequencing between step S313-S316, wherein, step S313 performs after step S309, and step S314 performs after step S310, and step S315 performs after step S311, and step S316 performs after step S312.
But, under normal circumstances, the metadata of disappearance can not be got in time from background server, therefore, as described in the background art, the metadata 1 write node 5 was once demoted at background server and metadata 2 repair correct before, if the metadata again on mirror-write node 5, because the metadata in the metadata on node 5 and other node is inconsistent, therefore will cause continuing degradation and write, and generate new degradation and write daily record.When the mirror-write operating frequency of metadata is higher, a large amount of degradation of generation is write daily record, thus adds the task amount of the metadata that background server will recover.The daily record loss write if demoted or daily record also do not perform, and at this time have other nodes break down again, distributed storage metadata may be caused to lose, and then lose user file.
For solving the problem, the embodiment of the present invention provides a kind of restorative procedure of distributed storage metadata, as shown in Figure 4, comprising:
S401, node 3 send message 3 to node 1, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S402, node 3 send message 3 to node 2, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S403, node 3 send message 3 to node 4, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
S404, node 3 send message 3 to node 5, and described message 3 carries metadata 2, and carry out the operation requests of mirror-write operation to metadata 2.
The described message 3 that S405, node 1 receiving node 3 send.
The described message 3 that S406, node 2 receiving node 3 send.
The described message 3 that S407, node 4 receiving node 3 send.
The described message 3 that S408, node 5 receiving node 3 send.
S409, node 1 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
S410, node 2 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
S411, node 4 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
If the metadata on S412 node 5 is imperfect, node 5 obtains the metadata of disappearance on node 5.
Owing to when there is node failure in node 5, mirror-write failure is carried out to metadata 1, therefore after node 5 receives the message 3 of node 3 transmission, by imperfect for the metadata object judged on this node, disappearance metadata 1, and then mirror-write operation cannot be carried out, in the embodiment of the present invention to metadata 2, now, node 5 obtains the metadata of disappearance on node 5.
In a kind of possible implementation, step S412 comprises:
If the metadata on S412a1 node 5 is imperfect, node 5 sends message 6 to node 3, and described message 6 is for the metadata of disappearance on requesting node 5.
The message 7 that S412a2, node 5 receiving node 3 send, described message 7 message carries the metadata of disappearance on node 5, obtains the metadata that node 5 lacks from described message 7.
In another kind of possible implementation, step S412 comprises:
If the metadata on S412b1 node 5 is imperfect, node 5 sends message 6 to MDS, and described message 6 is for the metadata of disappearance on requesting node 5.
S412b2, node 5 receive the message 7 that MDS sends, and described message 7 message carries the metadata of disappearance on node 5, obtain the metadata that node 5 lacks from described message 7.
Certainly, node 5 also can obtain the metadata of disappearance on node 5 by alternate manner, the embodiment of the present invention does not do concrete restriction to this.
S413, node 5, according to the metadata of the disappearance got, are repaired the metadata of disappearance on node 5.After metadata reparation on node 5 is complete, node 5 pairs of metadata 2 carry out mirror-write operation.
S414, node 5 send message 4 to node 3, and described message 4 is used to indicate metadata 2 mirror-write successful operation.
The described message 4 that S415, node 3 receiving node 1 send.
The described message 4 that S416, node 3 receiving node 2 send.
The described message 4 that S417, node 3 receiving node 4 send.
The described message 4 that S418, node 3 receiving node 5 send.
It should be noted that, in the restorative procedure of above-mentioned distributed meta data, inevitable execution sequencing is not had between step S401-S404, successively can perform any one step in step S401-S404, such as, perform step S401-S404 in turn, also can perform step S402, S403, S401, S404 etc. in turn, can also perform step S401-S404, the embodiment of the present invention does not do concrete restriction to this simultaneously.Do not have inevitable execution sequencing between step S405-S408, wherein, step S405 performs after step S401, and step S406 performs after step S402, and step S407 performs after step S403, and step S408 performs after step S404.Do not have inevitable execution sequencing between step S409-S412, wherein, step S409 performs after step S405, and step S410 performs after step S406, and step S411 performs after step S407, and step S412 performs after step S408.Do not have inevitable execution sequencing between step S415-S418, wherein, step S415 performs after step S509, and step S416 performs after the step s 410, and step S417 performs after step S411, and step S418 performs after step S414.
Based on the restorative procedure of the distributed storage metadata that the embodiment of the present invention provides, in the embodiment of the present invention, after node 5 receives the message 3 of carrying metadata 2 and metadata 2 being carried out to the operation requests of mirror-write operation that node 3 sends, if the metadata on node 5 is imperfect, node 5 obtains the metadata of disappearance on node 5, and then according to the metadata of the disappearance got, the metadata on node 5 is repaired, the metadata on node 5 is carried out reparation complete after, node 5 pairs of metadata 2 carry out mirror-write operation, and send message 4 to node 3, described message 4 is used to indicate metadata 2 mirror-write successful operation.That is, and unlike in prior art, when the metadata on node 5 is imperfect, need to wait for the reparation carrying out metadata after background server obtains reparation daily record, when the mirror-write operating frequency of metadata is higher, a large amount of degradation daily records may be generated.In the embodiment of the present invention, can metadata on node 5 imperfect time, in time the metadata on node 5 is repaired, makes this mirror-write successful operation, new degradation can not be generated and write daily record, and then follow-uply also can not generate more degradation and write daily record.Further, because the follow-up more degradation that can not generate writes daily record, therefore it also avoid in prior art when generate a large amount of degradations write daily record time, distributed storage metadata may be caused to lose, and then lose the problem of user file, add the reliability of metadata mirror image in distributed storage architecture.
For solving the problem, the embodiment of the present invention also provides a kind of restorative procedure of distributed storage metadata, is that executive agent is described, as shown in Figure 5, comprises with first node:
S501, first node receive the first message that Section Point sends, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Concrete, the first node in the embodiment of the present invention specifically refers to above-mentioned malfunctioning node, can be such as node 5; Section Point specifically refers to above-mentioned management node, can be such as node 3.
If the metadata on the described first node of S502 is imperfect, described first node obtains the metadata that described first node lacks.
In a kind of possible implementation, as shown in Figure 6, step S502 specifically can comprise:
If the metadata on the described first node of S502a1 is imperfect, described first node sends the 3rd message to described Section Point, and described 3rd message is for asking the metadata that described first node lacks.
S502a2, described first node receive the 4th message that described Section Point sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
In another kind of possible implementation, as shown in Figure 7, step S502 specifically can comprise:
If the metadata on the described first node of S502b1 is imperfect, described first node sends the 3rd message to MDS, and described 3rd message is for asking the metadata that described first node lacks.
S502b2, described first node receive the 4th message that described MDS sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
Certainly, if the metadata on described first node is imperfect, described first node can also obtain by alternate manner the metadata that described first node lacks, and the embodiment of the present invention does not do concrete restriction to this.
S503, described first node, according to the metadata of the disappearance got, are repaired the metadata on described first node.
S504, by described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
Based on the restorative procedure of the distributed storage metadata that the embodiment of the present invention provides, in the embodiment of the present invention, after first node receives first message of operation requests of carrying the first metadata and the first metadata being carried out to mirror-write operation that Section Point sends, when metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks, and then according to the metadata of the disappearance got, the metadata on described first node is repaired, by on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and send successful second message of the described first metadata mirror-write of instruction to Section Point.That is, and unlike in prior art, when the metadata on first node is imperfect, need to wait for the reparation carrying out metadata after background server obtains reparation daily record, when the mirror-write operating frequency of metadata is higher, a large amount of degradation daily records may be generated.In the embodiment of the present invention, can metadata on first node imperfect time, in time the metadata on first node is repaired, makes this mirror-write successful operation, can not generate new degradation and write daily record, and then the follow-up more degradation that also can not generate writes daily record.Further, because the follow-up more degradation that can not generate writes daily record, therefore it also avoid in prior art when generate a large amount of degradations write daily record time, distributed storage metadata may be caused to lose, and then lose the problem of user file, add the reliability of metadata mirror image in distributed storage architecture.
For solving the problem, the embodiment of the present invention also provides a kind of restorative procedure of distributed storage metadata, is that executive agent is described, as shown in Figure 8, comprises with Section Point:
S801, Section Point send the first message to first node, and described first message carries the first metadata, and carry out the operation requests of mirror-write operation to described first metadata.
Concrete, the first node in the embodiment of the present invention specifically refers to above-mentioned malfunctioning node, can be such as node 5; Section Point specifically refers to above-mentioned management node, can be such as node 3.
If the metadata on the described first node of S802 is imperfect, described Section Point receives the second message that described first node sends, and described second message is for asking the metadata that described first node lacks.
S803, described Section Point send the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
S804, described Section Point receive the 4th message that described first node sends, and described 4th message is used to indicate described first metadata mirror-write successful operation.
Further, as shown in Figure 9, before Section Point sends the first message (step S801) to first node, can also comprise:
S805, Section Point receive the first message that MDS sends.
S806, Section Point, according to described first message, carry out mirror-write operation to described first metadata.
Based on the restorative procedure of the distributed storage metadata that the embodiment of the present invention provides, in the embodiment of the present invention, at Section Point to after first node sends the first message of operation requests for carrying the first metadata and the first metadata being carried out to mirror-write operation, when metadata on described first node is imperfect, described Section Point receives the second message that described first node sends, and described second message is for asking the metadata that described first node lacks.And then described Section Point sends the 3rd message to described first node, and described 3rd message carries the metadata that described first node lacks.Like this, described first node can be repaired the metadata on described first node according to the metadata of the disappearance got, by on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and sends successful second message of the described first metadata mirror-write of instruction to Section Point.That is, and unlike in prior art, when the metadata on first node is imperfect, need to wait for the reparation carrying out metadata after background server obtains reparation daily record, when the mirror-write operating frequency of metadata is higher, a large amount of degradation daily records may be generated.In the embodiment of the present invention, can metadata on first node imperfect time, in time the metadata on first node is repaired, makes this mirror-write successful operation, can not generate new degradation and write daily record, and then the follow-up more degradation that also can not generate writes daily record.Further, because the follow-up more degradation that can not generate writes daily record, therefore it also avoid in prior art when generate a large amount of degradations write daily record time, distributed storage metadata may be caused to lose, and then lose the problem of user file, add the reliability of metadata mirror image in distributed storage architecture.
For solving the problem, the embodiment of the present invention also provides a kind of restorative procedure of distributed storage metadata, is that executive agent is described, as shown in Figure 10, comprises with MDS:
S1001, MDS send the first message to Section Point, and described first message carries the first metadata, and carry out the operation requests of mirror-write operation to described first metadata.
Concrete, the Section Point in the embodiment of the present invention specifically refers to above-mentioned management node, can be such as node 3.
If the metadata on the described first node of S1002 is imperfect, described MDS receives the second message that first node sends, and described second message is for asking the metadata that described first node lacks.
Concrete, the first node in the embodiment of the present invention specifically refers to above-mentioned malfunctioning node, can be such as node 5.
S1003, described MDS send the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Based on the restorative procedure of the distributed storage metadata that the embodiment of the present invention provides, in the embodiment of the present invention, at MDS to after Section Point sends first message of operation requests of carrying the first metadata and the first metadata being carried out to mirror-write operation, when metadata on described first node is imperfect, described MDS receives the second message that described first node sends, and described second message is for asking the metadata that described first node lacks.And then described MDS sends the 3rd message to described first node, and described 3rd message carries the metadata that described first node lacks.Like this, described first node can be repaired the metadata on described first node according to the metadata of the disappearance got, by on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and sends successful second message of the described first metadata mirror-write of instruction to Section Point.That is, and unlike in prior art, when the metadata on first node is imperfect, need to wait for the reparation carrying out metadata after background server obtains reparation daily record, when the mirror-write operating frequency of metadata is higher, a large amount of degradation daily records may be generated.In the embodiment of the present invention, can metadata on first node imperfect time, in time the metadata on first node is repaired, makes this mirror-write successful operation, can not generate new degradation and write daily record, and then the follow-up more degradation that also can not generate writes daily record.Further, because the follow-up more degradation that can not generate writes daily record, therefore it also avoid in prior art when generate a large amount of degradations write daily record time, distributed storage metadata may be caused to lose, and then lose the problem of user file, add the reliability of metadata mirror image in distributed storage architecture.
The embodiment of the present invention provides a kind of first node 110, and as shown in figure 11, described first node 110 comprises: receiving element 1101, processing unit 1102 and transmitting element.
Described receiving element 1101, for receiving the first message that Section Point sends, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described processing unit 1102, if imperfect for the metadata on described first node 110, described first node 110 obtains the metadata of disappearance on described first node 110.
Described processing unit 1102, also for the metadata according to the disappearance got, repairs the metadata on described first node 110.
Described processing unit 1102, also for by described first node 110 metadata repair complete after, mirror-write operation is carried out to described first metadata.
Described transmitting element, also for sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
In a kind of possible implementation, described processing unit 1102 specifically for:
Send the 3rd message by described transmitting element to described Section Point, described 3rd message is for asking the metadata of disappearance on described first node 110.
Received the 4th message of described Section Point transmission by described receiving element 1101, described 4th message carries the metadata of disappearance on described first node 110, obtains the metadata that described first node lacks from described 4th message.
In another kind of possible implementation, described processing unit 1102 specifically for:
Send the 3rd message by described transmitting element to MDS, described 3rd message is for asking the metadata of disappearance on described first node 110.
Received the 4th message of described MDS transmission by described receiving element 1101, described 4th message carries the metadata of disappearance on described first node 110, obtains the metadata that described first node lacks from described 4th message.
Concrete, the first node provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The first node provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The embodiment of the present invention provides a kind of Section Point 120, and as shown in figure 12, described Section Point 120 comprises: receiving element 1201 and transmitting element 1202.
Described transmitting element 1202, for sending the first message to first node, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described receiving element 1201, if the metadata on described first node is imperfect, receive the second message that described first node sends, described second message is for asking the metadata that described first node lacks.
Described transmitting element 1202, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Described receiving element 1201, also for receiving the 4th message that described first node sends, described 4th message is used to indicate described first metadata mirror-write successful operation.
Further, as shown in figure 13, described Section Point 120 also comprises processing unit 1203.
Described receiving element 1201, also for before sending the first message at described transmitting element 1202 to first node, receives described first message that meta data server MDS sends;
Described processing unit 1203, for according to described first message, carries out mirror-write operation to described first metadata.
Concrete, the Section Point provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The Section Point provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The embodiment of the present invention provides a kind of MDS140, and as shown in figure 14, described MDS140 comprises: receiving element 1401 transmitting element 1402.
Described transmitting element 1402, for sending the first message to Section Point, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described receiving element 1401, if the metadata on described first node is imperfect, receive the second message that first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitting element 1402, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Concrete, the MDS provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The MDS provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The embodiment of the present invention provides a kind of first node 150, and as shown in figure 15, described first node 150 comprises: receiver 1501, processor 1502, transmitter 1503 and bus 1504; Wherein, described receiver 1501, described processor 1502 and described transmitter 1503 are connected by described bus 1504 and complete mutual communication.
It should be noted that, processor 1502 here can be a processor, also can be the general designation of multiple treatment element.Such as, this processor 1502 can be that central processing unit is (English: CentralProcessingUnit, be called for short: CPU), also can be that specific integrated circuit is (English: ApplicationSpecificIntegratedCircuit, be called for short: ASIC), or be configured to the one or more integrated circuit implementing the embodiment of the present invention, such as: one or more microprocessor is (English: digitalsingnalprocessor, be called for short: DSP), or, one or more field programmable gate array is (English: FieldProgrammableGateArray, be called for short: FPGA).
Bus 1504 can be that industry standard architecture is (English: IndustryStandardArchitecture, be called for short: ISA) bus, peripheral component interconnect (English: PeripheralComponent, be called for short: PCI) (English: ExtendedIndustryStandardArchitecture is called for short: EISA) bus etc. for bus or extended industry-standard architecture.This bus 1504 can be divided into address bus, data bus, control bus etc.For ease of representing, only representing bus 1504 with a thick line in Figure 15, but not representing the bus only having a bus or a type.
Described receiver 1501, for receiving the first message that Section Point sends, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described processor 1502, if imperfect for the metadata on described first node 150, described first node 150 obtains the metadata of disappearance on described first node 150.
Described processor 1502, also for the metadata according to the disappearance got, repairs the metadata on described first node 150.
Described processor 1502, also for by described first node 150 metadata repair complete after, mirror-write operation is carried out to described first metadata.
Described transmitter 1503, also for sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
In a kind of possible implementation, described processor 1502 specifically for:
Send the 3rd message by described transmitter 1503 to described Section Point, described 3rd message is for asking the metadata of disappearance on described first node 150.
Received the 4th message of described Section Point transmission by described receiver 1501, described 4th message carries the metadata of disappearance on described first node 150, obtains the metadata that described first node lacks from described 4th message.
In another kind of possible implementation, described processor 1502 specifically for:
Send the 3rd message by described transmitter 1503 to MDS, described 3rd message is for asking the metadata of disappearance on described first node 150.
Received the 4th message of described MDS transmission by described receiver 1501, described 4th message carries the metadata of disappearance on described first node 150, obtains the metadata that described first node lacks from described 4th message.
Concrete, the first node provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The first node provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The embodiment of the present invention provides a kind of Section Point 160, and as shown in figure 16, described Section Point 160 comprises: receiver 1601, transmitter 1602 and bus 1603; Wherein, described receiver 1601 and described transmitter 1602 are connected by described bus 1603 and complete mutual communication.
It should be noted that, bus 1603 here can be isa bus, pci bus or eisa bus etc.This bus 1603 can be divided into address bus, data bus, control bus etc.For ease of representing, only representing bus 1603 with a thick line in Figure 16, but not representing the bus only having a bus or a type.
Described transmitter 1602, for sending the first message to first node, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described receiver 1601, if the metadata on described first node is imperfect, receive the second message that described first node sends, described second message is for asking the metadata that described first node lacks.
Described transmitter 1602, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Described receiver 1601, also for receiving the 4th message that described first node sends, described 4th message is used to indicate described first metadata mirror-write successful operation.
Further, as shown in figure 17, described Section Point 160 also comprises processor 1604.
Described receiver 1601, also for before sending the first message at described transmitter 1602 to first node, receives described first message that MDS sends.
Described processor 1604, for according to described first message, carries out mirror-write operation to described first metadata.
It should be noted that, processor 1604 here can be a processor, also can be the general designation of multiple treatment element.Such as, this processor 1604 can be CPU, can be also ASIC, or is configured to the one or more integrated circuit implementing the embodiment of the present invention, such as: one or more DSP, or, one or more FPGA.
Concrete, the Section Point provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The Section Point provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The embodiment of the present invention provides a kind of MDS180, and as shown in figure 18, described MDS180 comprises: receiver 1801, transmitter 1802 and bus 1803; Wherein, described receiver 1801 and described transmitter 1802 are connected by described bus 1803 and complete mutual communication.
It should be noted that, bus 1803 here can be isa bus, pci bus or eisa bus etc.This bus 1803 can be divided into address bus, data bus, control bus etc.For ease of representing, only representing bus 1803 with a thick line in Figure 18, but not representing the bus only having a bus or a type.
Described transmitter 1802, for sending the first message to Section Point, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata.
Described receiver 1801, if the metadata on described first node is imperfect, receive the second message that first node sends, described second message is for asking the metadata that described first node lacks.
Described transmitter 1802, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
Concrete, the MDS provided by the embodiment of the present invention carries out the method for distributed storage metadata reparation can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
The MDS provided due to the embodiment of the present invention can perform the restorative procedure of above-mentioned distributed storage metadata, and therefore its obtainable technique effect of institute can with reference to said method embodiment, and the embodiment of the present invention does not repeat them here.
In addition, the embodiment of the present invention also provides a kind of and calculates readable media (or medium), is included in the computer-readable instruction carrying out following operation when being performed: perform the operation in the embodiment of the method in above-described embodiment as shown in Fig. 4-Figure 10.
In addition, the embodiment of the present invention also provides a kind of computer program, comprises above computer computer-readable recording medium.
Should understand, in various embodiments of the present invention, the size of the sequence number of above-mentioned each process does not also mean that the priority of execution sequence, and the execution sequence of each process should be determined with its function and internal logic, and should not form any restriction to the implementation process of the embodiment of the present invention.
Those skilled in the art can be well understood to, for convenience and simplicity of description, the device of foregoing description, only be illustrated with the division of above-mentioned each functional module, in practical application, can distribute as required and by above-mentioned functions and be completed by different functional modules, the inner structure by device is divided into different functional modules, to complete all or part of function described above.The specific works process of the system of foregoing description, device and unit, with reference to the corresponding process in preceding method embodiment, can not repeat them here.
In several embodiments that the application provides, should be understood that, disclosed system, apparatus and method, can realize by another way.Such as, device embodiment described above is only schematic, such as, the division of described module or unit, be only a kind of logic function to divide, actual can have other dividing mode when realizing, such as multiple unit or assembly can in conjunction with or another system can be integrated into, or some features can be ignored, or do not perform.Another point, shown or discussed coupling each other or direct-coupling or communication connection can be by some interfaces, and the indirect coupling of device or unit or communication connection can be electrical, machinery or other form.
The described unit illustrated as separating component or can may not be and physically separates, and the parts as unit display can be or may not be physical location, namely can be positioned at a place, or also can be distributed in multiple network element.Some or all of unit wherein can be selected according to the actual needs to realize the object of the present embodiment scheme.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, also can be that the independent physics of unit exists, also can two or more unit in a unit integrated.Above-mentioned integrated unit both can adopt the form of hardware to realize, and the form of SFU software functional unit also can be adopted to realize.
If described integrated unit using the form of SFU software functional unit realize and as independently production marketing or use time, can be stored in a computer read/write memory medium.Based on such understanding, the part that technical scheme of the present invention contributes to prior art in essence in other words or all or part of of this technical scheme can embody with the form of software product, this computer software product is stored in a storage medium, comprising some instructions in order to make a computer equipment (can be personal computer, server, or the network equipment etc.) or processor (processor) perform all or part of step of method described in each embodiment of the present invention.And aforesaid storage medium comprises: USB flash disk, portable hard drive, ROM, RAM), magnetic disc or CD etc. various can be program code stored medium.
The above; be only the specific embodiment of the present invention, but protection scope of the present invention is not limited thereto, is anyly familiar with those skilled in the art in the technical scope that the present invention discloses; change can be expected easily or replace, all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection domain of described claim.

Claims (13)

1. a restorative procedure for distributed storage metadata, is characterized in that, described method comprises:
First node receives the first message that Section Point sends, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks;
Described first node, according to the metadata of the disappearance got, is repaired the metadata on described first node;
By on described first node metadata repair complete after, described first node carries out mirror-write operation to described first metadata, and sends the second message to described Section Point, and described second message is used to indicate described first metadata mirror-write successful operation.
2. method according to claim 1, is characterized in that, described first node obtains the metadata that described first node lacks, and comprising:
Described first node sends the 3rd message to described Section Point, and described 3rd message is for asking the metadata that described first node lacks;
Described first node receives the 4th message that described Section Point sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
3. method according to claim 1, is characterized in that, described first node obtains the metadata that described first node lacks, and comprising:
Described first node sends the 3rd message to meta data server MDS, and described 3rd message is for asking the metadata that described first node lacks;
Described first node receives the 4th message that described MDS sends, and described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
4. a restorative procedure for distributed storage metadata, is characterized in that, described method comprises:
Section Point sends the first message to first node, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described Section Point receives the second message that described first node sends, and described second message is for asking the metadata that described first node lacks;
Described Section Point sends the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node;
Described Section Point receives the 4th message that described first node sends, and described 4th message is used to indicate described first metadata mirror-write successful operation.
5. method according to claim 4, is characterized in that, before described Section Point sends the first message to first node, also comprises:
Described Section Point receives described first message that meta data server MDS sends;
Described Section Point, according to described first message, carries out mirror-write operation to described first metadata.
6. a restorative procedure for distributed storage metadata, is characterized in that, described method comprises:
Meta data server MDS sends the first message to Section Point, and described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
If the metadata on described first node is imperfect, described MDS receives the second message that first node sends, and described second message is for asking the metadata that described first node lacks;
Described MDS sends the 3rd message to described first node, and described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
7. a first node, is characterized in that, described first node comprises: receiving element, processing unit and transmitting element;
Described receiving element, for receiving the first message that Section Point sends, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described processing unit, if the metadata on described first node is imperfect, described first node obtains the metadata that described first node lacks;
Described processing unit, also for the metadata according to the disappearance got, repairs the metadata on described first node;
Described processing unit, also for by described first node metadata repair complete after, mirror-write operation is carried out to described first metadata;
Described transmitting element, also for sending the second message to described Section Point, described second message is used to indicate described first metadata mirror-write successful operation.
8. first node according to claim 7, is characterized in that, described processing unit specifically for:
Send the 3rd message by described transmitting element to described Section Point, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described Section Point transmission by described receiving element, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
9. first node according to claim 7, is characterized in that, described processing unit specifically for:
Send the 3rd message by described transmitting element to meta data server MDS, described 3rd message is for asking the metadata that described first node lacks;
Received the 4th message of described MDS transmission by described receiving element, described 4th message carries the metadata that described first node lacks, and obtains the metadata that described first node lacks from described 4th message.
10. a Section Point, is characterized in that, described Section Point comprises: receiving element and transmitting element;
Described transmitting element, for sending the first message to first node, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiving element, if the metadata on described first node is imperfect, receive the second message that described first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitting element, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node;
Described receiving element, also for receiving the 4th message that described first node sends, described 4th message is used to indicate described first metadata mirror-write successful operation.
11. Section Points according to claim 10, is characterized in that, described Section Point also comprises processing unit;
Described receiving element, also for before sending the first message at described transmitting element to first node, receives described first message that meta data server MDS sends;
Described processing unit, for according to described first message, carries out mirror-write operation to described first metadata.
12. 1 kinds of meta data server MDS, is characterized in that, described MDS comprises: receiving element transmitting element;
Described transmitting element, for sending the first message to Section Point, described first message carries the first metadata, and carries out the operation requests of mirror-write operation to described first metadata;
Described receiving element, if the metadata on described first node is imperfect, receive the second message that first node sends, described second message is for asking the metadata that described first node lacks;
Described transmitting element, also for sending the 3rd message to described first node, described 3rd message carries the metadata that described first node lacks, and is repaired according to the metadata that described first node lacks by described first node to the metadata on described first node.
13. 1 kinds of distributed storage architecture, it is characterized in that, described distributed storage architecture comprises the first node as described in any one of claim 7-9, the Section Point as described in claim 10 or 11 and meta data server MDS as claimed in claim 12.
CN201510799090.XA 2015-11-18 2015-11-18 Restorative procedure, the apparatus and system of distributed storage metadata Active CN105373449B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510799090.XA CN105373449B (en) 2015-11-18 2015-11-18 Restorative procedure, the apparatus and system of distributed storage metadata

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510799090.XA CN105373449B (en) 2015-11-18 2015-11-18 Restorative procedure, the apparatus and system of distributed storage metadata

Publications (2)

Publication Number Publication Date
CN105373449A true CN105373449A (en) 2016-03-02
CN105373449B CN105373449B (en) 2019-06-21

Family

ID=55375670

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510799090.XA Active CN105373449B (en) 2015-11-18 2015-11-18 Restorative procedure, the apparatus and system of distributed storage metadata

Country Status (1)

Country Link
CN (1) CN105373449B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106685744A (en) * 2017-03-13 2017-05-17 福建中金在线信息科技有限公司 Fault elimination method, apparatus and system
CN107153508A (en) * 2016-03-03 2017-09-12 杭州海康威视数字技术股份有限公司 A kind of date storage method and device
CN107861842A (en) * 2017-11-08 2018-03-30 郑州云海信息技术有限公司 A kind of metadata damage detection method, system, equipment and storage medium
WO2021005444A1 (en) * 2019-07-08 2021-01-14 International Business Machines Corporation Mirroring write operations across data storage devices

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103744745A (en) * 2013-12-13 2014-04-23 北京奇虎科技有限公司 Data storage detection method and device and distributed storage system
CN104462432A (en) * 2014-12-15 2015-03-25 成都英力拓信息技术有限公司 Self-adaptive distributed computing method
CN104679893A (en) * 2015-03-18 2015-06-03 成都金本华科技股份有限公司 Information retrieval method based on big data

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103744745A (en) * 2013-12-13 2014-04-23 北京奇虎科技有限公司 Data storage detection method and device and distributed storage system
CN104462432A (en) * 2014-12-15 2015-03-25 成都英力拓信息技术有限公司 Self-adaptive distributed computing method
CN104679893A (en) * 2015-03-18 2015-06-03 成都金本华科技股份有限公司 Information retrieval method based on big data

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107153508A (en) * 2016-03-03 2017-09-12 杭州海康威视数字技术股份有限公司 A kind of date storage method and device
CN107153508B (en) * 2016-03-03 2020-07-24 杭州海康威视数字技术股份有限公司 Data storage method and device
CN106685744A (en) * 2017-03-13 2017-05-17 福建中金在线信息科技有限公司 Fault elimination method, apparatus and system
CN107861842A (en) * 2017-11-08 2018-03-30 郑州云海信息技术有限公司 A kind of metadata damage detection method, system, equipment and storage medium
CN107861842B (en) * 2017-11-08 2021-10-15 郑州云海信息技术有限公司 Metadata damage detection method, system, equipment and storage medium
WO2021005444A1 (en) * 2019-07-08 2021-01-14 International Business Machines Corporation Mirroring write operations across data storage devices
US10922009B2 (en) 2019-07-08 2021-02-16 International Business Machines Corporation Mirroring write operations across data storage devices
GB2600298A (en) * 2019-07-08 2022-04-27 Ibm Mirroring write operations across data storage devices
GB2600298B (en) * 2019-07-08 2023-03-01 Ibm Mirroring write operations across data storage devices

Also Published As

Publication number Publication date
CN105373449B (en) 2019-06-21

Similar Documents

Publication Publication Date Title
US10379951B2 (en) Hierarchic storage policy for distributed object storage systems
US10725692B2 (en) Data storage method and apparatus
US9715436B2 (en) System and method for managing raid storage system having a hot spare drive
CN103268201B (en) A kind of date storage method, memory storage and read method
CN106463179A (en) Method, apparatus and system for handling data error events with memory controller
KR101572038B1 (en) Correction of errors in a memory array
CN105373449A (en) Distributed storage metadata recovery method, apparatus and system
US10949314B2 (en) Method and apparatus for failure recovery of storage device
CN106557266B (en) Method and apparatus for redundant array of independent disks RAID
US10860447B2 (en) Database cluster architecture based on dual port solid state disk
CN105653345A (en) Method and device supporting data nonvolatile random access
CN105159800A (en) Data recovery method, data backup method, data recovery equipment, data backup equipment, data recovery system and data backup system
CN110442298B (en) Storage equipment abnormality detection method and device and distributed storage system
CN105550076A (en) Image splicing control system and redundant hot backup method thereof
US9037948B2 (en) Error correction for memory systems
CN102760109A (en) Data communication method, device and system
CN113448844B (en) Method and device for regression testing and electronic equipment
CN104035855A (en) Method and device for monitoring hard disks
CN104503868A (en) Data synchronizing method, device and system
CN111488124A (en) Data updating method and device, electronic equipment and storage medium
CN115202589A (en) Placement group member selection method, device, equipment and readable storage medium
CN105159790A (en) Data rescue method and file server
CN106844094B (en) File repair method and device
CN106940684B (en) Method and device for writing data according to bits
CN110968456A (en) Method and device for processing fault disk in distributed storage system

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