CN109729014A - A kind of message storage method and device - Google Patents
A kind of message storage method and device Download PDFInfo
- Publication number
- CN109729014A CN109729014A CN201711046599.2A CN201711046599A CN109729014A CN 109729014 A CN109729014 A CN 109729014A CN 201711046599 A CN201711046599 A CN 201711046599A CN 109729014 A CN109729014 A CN 109729014A
- Authority
- CN
- China
- Prior art keywords
- message
- memory
- peripheral
- peripheral memory
- internal
- 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
Abstract
The application proposes a kind of message storage method and device, it is related to data communication field, the message storage method, comprising: when the message of internal memory accumulation capacity meets preset condition, the message cached in the internal memory is moved into peripheral memory sequentially in time；During moving message, when peripheral memory back-pressure moves access the case where, the message moved to peripheral memory is abandoned.The internal memory that big bandwidth is used using high priority message can be effectively ensured, the message of low priority is shifted to peripheral memory in congestion, in storage link according to height prioritized operation, efficiently use the bandwidth of internal storage and the capacity of peripheral memory.
Description
Technical field
The present invention relates to data communication fields, and in particular to a kind of message storage method and device.
Background technique
With the continuous development of network, the bandwidth demand of service provider and data center is at an unprecedented rate
Enhancing.It is faced with new demand and challenge, new equipment quickly must be developed and be listed, and require reducing Total Cost Ownership
While keep high availability.
But for core backbone, the bandwidth development of memory technology meets with bottleneck, DDR (Double Data at present
Rate, Double Data Rate synchronous DRAM), QDR (Quad Data Rate, Quad Data Rate synchronous dynamic random storage
Device) and newest HBM (High Bandwidth Memory, high bandwidth video memory) and HMC (Hybrid Memory Cube,
Mixing memory cube) all it is unable to satisfy the memory bandwidth requirement of Tbps rank on core backbone.
The status that bandwidth requirement cannot be met in face of currently stored device generallys use the technology of inside and outside collaboration caching to mention
High bandwidth and memory capacity.When external memory bandwidth deficiency, majority design uses the strategy for doing bandwidth allocation in advance, but time
After delay is big, and caching is used up, the problem of non-congested high priority message may also be dropped.
Summary of the invention
The present invention provides a kind of message storage method and device, is realized by abandoning segment message automatically and improves memory bandwidth
Effect.
In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is as follows:
In a first aspect, the present invention provides a kind of message storage method, comprising:
When the message of internal memory accumulation capacity meets preset condition, the message that will be cached in the internal memory
Peripheral memory is moved sequentially in time；
During moving message, when peripheral memory back-pressure moves access the case where, abandons and removed to peripheral memory
The message of shifting.
Preferably, the method also includes:
When needing to dispatch outgoing message, the position of packet storage to be dispatched is judged；
When the packet storage of scheduling is in the internal storage, the message of the scheduling is gone out from the internal storage
Team；
When the packet storage of scheduling is in the peripheral memory, the message of the scheduling is gone out from the peripheral memory
Team；
Preferably, the preset condition includes:
Predetermined threshold value.
Preferably, abandoning the message moved to peripheral memory includes:
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal, until
The case where peripheral memory back-pressure moves access disappearance.
Second aspect, the present invention also provides a kind of packet storage devices, comprising:
Moving module is set as built-in depositing when the message of internal memory accumulation capacity meets preset condition by described
The message cached in reservoir moves peripheral memory sequentially in time；
Discard module is set as during moving message, when peripheral memory back-pressure moves access the case where, is abandoned
The message moved to peripheral memory.
Preferably, the packet storage device further include:
Judgment module is set as when needing to dispatch outgoing message, judges the position of packet storage to be dispatched；
When the packet storage of scheduling is in the internal storage, the message of the scheduling is gone out from the internal storage
Team；
When the packet storage of scheduling is in the peripheral memory, the message of the scheduling is gone out from the peripheral memory
Team；
Preferably, the preset condition of the moving module includes:
Predetermined threshold value.
Preferably, the discard module abandons the message moved to peripheral memory and includes:
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal, until
The case where peripheral memory back-pressure moves access disappearance.
The third aspect, the present invention also provides a kind of packet storage devices, comprising: memory and processor；
The memory, for saving executable instruction；
The processor, the executable instruction saved for executing the memory, proceeds as follows:
When the message of internal memory accumulation capacity meets preset condition, the message that will be cached in the internal memory
Peripheral memory is moved sequentially in time；
During moving message, when peripheral memory back-pressure moves access the case where, abandons and removed to peripheral memory
The message of shifting.
Compared to the prior art the present invention, has the following beneficial effects:
The message of congestion message or low priority is moved peripheral memory by technical solution of the present invention；When moving band
When wide insufficient, segment message is abandoned to guarantee global storage bandwidth, can be effectively ensured and be used big band using high priority message
The message of wide internal memory, low priority is shifted to peripheral memory in congestion, excellent according to height in storage link
First grade operation, efficiently uses the bandwidth of internal storage and the capacity of peripheral memory.
Meanwhile being encountered in the insufficient situation of bandwidth if moving, message can also be moved by discarding low priority, be protected
The message of card high priority continues to use internal storage.
Detailed description of the invention
Fig. 1 is the flow chart of the message storage method of the embodiment of the present invention；
Fig. 2 is the structural schematic diagram of the packet storage device of the embodiment of the present invention；
Fig. 3 is the schematic diagram of the realization dropping packets of the embodiment of the present invention.
Specific embodiment
To keep goal of the invention of the invention, technical scheme and beneficial effects more clear, with reference to the accompanying drawing to this
The embodiment of invention is illustrated, it should be noted that in the absence of conflict, in the embodiment and embodiment in the application
Feature can mutual any combination.
As shown in Figure 1, the embodiment of the present invention provides a kind of message storage method, comprising:
S101, when internal memory message accumulation capacity meet preset condition when, will be cached in the internal memory
Message move peripheral memory sequentially in time；
S102, during moving message, when peripheral memory back-pressure moves access the case where, abandon and stored to peripheral hardware
The message that device is moved.
In the embodiment of the present invention, internal memory generate message accumulation, when being stacked into preset condition, then according to message into
The time order and function relationship for entering internal memory will enter at first the message of internal memory, move the peripheral memory of low speed
In.Give big bandwidth to non-congested message.
Because the bandwidth of peripheral memory is limited, during moving message, if there is processing bottleneck in peripheral memory,
Because of its bandwidth ability deficiency, back-pressure moves access, at this time, it is desirable to which operation is moved in internal memory stopping, if pause is moved
Data manipulation continues to accumulate it will cause internal memory, or even influences whether the message of high priority, therefore, current to need
The message moved actively is abandoned.
The method also includes:
When outgoing message to be dispatched, the position of the packet storage to be dispatched described in judgement；
When the packet storage of scheduling is in the internal storage, by the message of the scheduling from the internal storage tune
Degree output；
When the packet storage of scheduling is in the peripheral memory, by the message of the scheduling from the peripheral memory tune
Degree output；
In the embodiment of the present invention, when authorization issues, when dispatching out team, first determine whether that message is stored in built-in storage
Device or peripheral memory are directly dispatched from internal memory and are exported if being stored in internal memory, if being stored in outer
If memory, then dispatches and export from peripheral memory.
The preset condition includes:
Predetermined threshold value.
In step S102, abandoning the message moved to peripheral memory includes:
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal.
Specifically, abandoning the segment message moved to peripheral memory includes:
When peripheral memory back-pressure occur and moving access, the segment message moved to peripheral memory is abandoned, until outer
If the case where memory back-pressure moves access disappearance.
As shown in Fig. 2, the embodiment of the present invention also provides a kind of packet storage device, comprising:
Moving module is set as built-in depositing when the message of internal memory accumulation capacity meets preset condition by described
The message cached in reservoir moves peripheral memory sequentially in time；
Discard module is set as during moving message, when peripheral memory back-pressure moves access the case where, is abandoned
The message moved to peripheral memory.
Preferably, the packet storage device further include:
Judgment module is set as when needing to dispatch outgoing message, judges the position of packet storage to be dispatched；
When the packet storage of scheduling is in the internal storage, the message of the scheduling is gone out from the internal storage
Team；
When the packet storage of scheduling is in the peripheral memory, the message of the scheduling is gone out from the peripheral memory
Team；
The preset condition of the preferably described moving module includes:
Predetermined threshold value.
The preferably described discard module abandons the message moved to peripheral memory
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal, until
The case where peripheral memory back-pressure moves access disappearance.
The embodiment of the present invention also provides a kind of packet storage device, comprising: memory and processor；
The memory, for saving executable instruction；
The processor, the executable instruction saved for executing the memory, proceeds as follows:
When the message of internal memory accumulation capacity meets preset condition, the message that will be cached in the internal memory
Peripheral memory is moved sequentially in time；
During moving message, when peripheral memory back-pressure moves access the case where, abandons and removed to peripheral memory
The message of shifting.
Embodiment 1
Internal memory has with roomy, the small feature of capacity.Correspondingly, peripheral memory bandwidth is small, capacity is big.Such as figure
Shown in 3, in the present invention, the message of all arrivals does not go directly to peripheral memory in storage to internal memory
(as shown in data flow 1 in Fig. 3).
If authorization issues in time, internal memory does not generate message accumulation, then message will begin before scheduled team out
It stays in internal memory eventually, when needing schedules message to go out team, then directly goes out team's (data flow 2 in such as Fig. 3 from internal memory
It is shown), bandwidth is sufficiently large.
If authorization issues not in time, internal memory generates message accumulation, and when being stacked into certain thresholding, (this thresholding can appoint
Meaning configuration), then enter the time order and function relationship of internal memory according to message, the message of internal memory will be entered at first, removed
It moves on in peripheral memory and goes (as shown in data flow 3 in Fig. 3), in this way, the large capacity feature of peripheral memory can then be answered
It uses on the message of low priority.When to dispatch out team, first determine whether that message is stored in internal memory or peripheral hardware
Memory goes out team from peripheral memory if being stored in peripheral memory (as shown in data flow 4 in Fig. 3).
When moving generation, allochthonous message is the message of earliest deposit internal memory, also means that this number
It is congestion data stream or low priority message according to message, cannot slowly authorizes, to has been shifted to the peripheral hardware storage of low speed
Big bandwidth is given the non-congested message of high priority by device.And high prior message will not be shifted to peripheral memory,
All go out team in time from internal memory after to authorization.
It, can be anti-because its bandwidth ability is insufficient during moving message because the bandwidth of peripheral memory is limited
Pressure moves access.If suspending moving data operation during back-pressure, continues to accumulate it will cause internal memory, finally will affect
To the message of high priority, need actively to abandon part of message to solve congestion problems, otherwise the message of high priority enters
Team goes out team also and will receive influence.At this point, being moved low priority message using discarding to delay in order to not influence high priority message
Solution moves bandwidth deficiency.To guarantee that internal memory does not continue to accumulate, guarantee that high priority message is still smoothly joined the team
Team out.
Whether discard is moved in unlatching, can be made configurable.
Although disclosed embodiment is as above, its content is only to facilitate understand technical side of the invention
Case and the embodiment used, are not intended to limit the present invention.Any those skilled in the art to which this invention pertains, not
Under the premise of being detached from disclosed core technology scheme, any modification and change can be made in form and details in implementation
Change, but protection scope defined by the present invention, the range that the appended claims that must still be subject to limits.
Claims (9)
1. a kind of message storage method characterized by comprising
When internal memory message accumulation capacity meet preset condition when, by the message cached in the internal memory according to
Time sequencing moves peripheral memory；
During moving message, when peripheral memory back-pressure moves access the case where, abandons and moved to peripheral memory
Message.
2. message storage method as described in claim 1, it is characterised in that: the method also includes:
When needing to dispatch outgoing message, the position of packet storage to be dispatched is judged；
When the packet storage of scheduling is in the internal storage, the message of the scheduling is gone out into team from the internal storage；
When the packet storage of scheduling is in the peripheral memory, the message of the scheduling is gone out into team from the peripheral memory.
3. message storage method as described in claim 1, it is characterised in that: the preset condition includes:
Predetermined threshold value.
4. message storage method as described in claim 1, it is characterised in that: abandon the message packet moved to peripheral memory
It includes:
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal, until peripheral hardware
The case where memory back-pressure moves access disappearance.
5. a kind of packet storage device, it is characterised in that: include:
Moving module is set as when the message of internal memory accumulation capacity meets preset condition, by the internal memory
The message of middle caching moves peripheral memory sequentially in time；
Discard module is set as during moving message, when peripheral memory back-pressure moves access the case where, is abandoned outside
If the message that memory is moved.
6. packet storage device as claimed in claim 5, it is characterised in that: further include:
Judgment module is set as when needing to dispatch outgoing message, judges the position of packet storage to be dispatched；
When the packet storage of scheduling is in the internal storage, the message of the scheduling is gone out into team from the internal storage；
When the packet storage of scheduling is in the peripheral memory, the message of the scheduling is gone out into team from the peripheral memory.
7. packet storage device as claimed in claim 5, it is characterised in that: the preset condition of the moving module includes:
Predetermined threshold value.
8. packet storage device as claimed in claim 5, it is characterised in that: the discard module is abandoned removes to peripheral memory
The message of shifting includes:
Trigger signal is received, all or part of message moved to peripheral memory is abandoned according to trigger signal, until peripheral hardware
The case where memory back-pressure moves access disappearance.
9. a kind of packet storage device characterized by comprising memory and processor；
The memory, for saving executable instruction；
The processor, the executable instruction saved for executing the memory, proceeds as follows:
When internal memory message accumulation capacity meet preset condition when, by the message cached in the internal memory according to
Time sequencing moves peripheral memory；
During moving message, when peripheral memory back-pressure moves access the case where, abandons and moved to peripheral memory
Message.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711046599.2A CN109729014B (en) | 2017-10-31 | 2017-10-31 | Message storage method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711046599.2A CN109729014B (en) | 2017-10-31 | 2017-10-31 | Message storage method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109729014A true CN109729014A (en) | 2019-05-07 |
| CN109729014B CN109729014B (en) | 2023-09-12 |
Family
ID=66294368
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711046599.2A Active CN109729014B (en) | 2017-10-31 | 2017-10-31 | Message storage method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109729014B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112804156A (en) * | 2019-11-13 | 2021-05-14 | 深圳市中兴微电子技术有限公司 | Congestion avoidance method and device and computer readable storage medium |
| WO2021259321A1 (en) * | 2020-06-23 | 2021-12-30 | 中兴通讯股份有限公司 | Storage scheduling method, device, and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101784082A (en) * | 2009-12-22 | 2010-07-21 | 中兴通讯股份有限公司 | Method and device for enhancing service quality in wireless local area network |
| CN102571535A (en) * | 2010-12-22 | 2012-07-11 | 深圳市恒扬科技有限公司 | Device and method for delaying data and communication system |
| CN102594691A (en) * | 2012-02-23 | 2012-07-18 | 中兴通讯股份有限公司 | Method and device for processing message |
| CN103888377A (en) * | 2014-03-28 | 2014-06-25 | 华为技术有限公司 | Message cache method and device |
| WO2017028545A1 (en) * | 2015-08-14 | 2017-02-23 | 中兴通讯股份有限公司 | Port queue congestion monitoring method and system |
| CN107222427A (en) * | 2016-03-22 | 2017-09-29 | 华为技术有限公司 | The method and relevant device of a kind of Message processing |
-
2017
- 2017-10-31 CN CN201711046599.2A patent/CN109729014B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101784082A (en) * | 2009-12-22 | 2010-07-21 | 中兴通讯股份有限公司 | Method and device for enhancing service quality in wireless local area network |
| CN102571535A (en) * | 2010-12-22 | 2012-07-11 | 深圳市恒扬科技有限公司 | Device and method for delaying data and communication system |
| CN102594691A (en) * | 2012-02-23 | 2012-07-18 | 中兴通讯股份有限公司 | Method and device for processing message |
| CN103888377A (en) * | 2014-03-28 | 2014-06-25 | 华为技术有限公司 | Message cache method and device |
| WO2017028545A1 (en) * | 2015-08-14 | 2017-02-23 | 中兴通讯股份有限公司 | Port queue congestion monitoring method and system |
| CN107222427A (en) * | 2016-03-22 | 2017-09-29 | 华为技术有限公司 | The method and relevant device of a kind of Message processing |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112804156A (en) * | 2019-11-13 | 2021-05-14 | 深圳市中兴微电子技术有限公司 | Congestion avoidance method and device and computer readable storage medium |
| WO2021259321A1 (en) * | 2020-06-23 | 2021-12-30 | 中兴通讯股份有限公司 | Storage scheduling method, device, and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109729014B (en) | 2023-09-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102999386B (en) | Method for achieving multi-message queue mechanism to improve performance in single process | |
| EP2702730B1 (en) | Effective circuits in packet-switched networks | |
| CN102915254B (en) | task management method and device | |
| CN101557348B (en) | Message forwarding method and device based on token bucket | |
| CN102594691B (en) | A kind of method and device handling message | |
| CN109697122A (en) | Task processing method, equipment and computer storage medium | |
| US9774651B2 (en) | Method and apparatus for rapid data distribution | |
| KR20140100529A (en) | Providing path information in mixed communication networks | |
| CN109729014A (en) | A kind of message storage method and device | |
| CN102811176B (en) | A kind of data flow control method and device | |
| CN107430496A (en) | Frame is abandoned for long-range display | |
| US8503472B2 (en) | Partial bandwidth request techniques in wireless networks | |
| CN103617132B (en) | A kind of ethernet terminal based on shared storage sends implementation method and terminal installation | |
| CN102521159A (en) | Method for dynamically scheduling packet data of on-chip and off-chip memories according to traffic | |
| CN105764091A (en) | Method of effectively utilizing wireless bandwidth | |
| CN102883420A (en) | Transmission scheduling method and device under sleeping mode | |
| CN109039934B (en) | Space DTN network congestion control method and system | |
| CN102404225B (en) | Method for rapid enqueue of packet for differential queue service system | |
| CN102546423A (en) | Method and device for queue scheduling and network device | |
| US9891840B2 (en) | Method and arrangement for controlling requests to a shared electronic resource | |
| CN105337888B (en) | Load-balancing method, device and virtual switch based on multicore forwarding | |
| CN111385223B (en) | Flow control method, device, system and storage medium | |
| CN103389950B (en) | Anti-jamming multichannel data transmission method based on capacity prediction | |
| CN111400028B (en) | Load balancing processing method for train management | |
| CN116889024A (en) | Data stream transmission method, device and network equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |