CN107239320A - The method of process status in real-time preservation client computer based on virtualization technology - Google Patents
The method of process status in real-time preservation client computer based on virtualization technology Download PDFInfo
- Publication number
- CN107239320A CN107239320A CN201710233176.5A CN201710233176A CN107239320A CN 107239320 A CN107239320 A CN 107239320A CN 201710233176 A CN201710233176 A CN 201710233176A CN 107239320 A CN107239320 A CN 107239320A
- Authority
- CN
- China
- Prior art keywords
- client computer
- absorbed
- virtualization
- backup
- client
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/14—Error detection or correction of the data by redundancy in operation
- G06F11/1402—Saving, restoring, recovering or retrying
- G06F11/1446—Point-in-time backing up or restoration of persistent data
- G06F11/1458—Management of the backup or restore process
- G06F11/1469—Backup restoration techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Retry When Errors Occur (AREA)
- Hardware Redundancy (AREA)
Abstract
The present invention relates to a kind of method of process status in real-time preservation client computer based on virtualization technology.This method includes：1) PID of specific process in client computer is notified to give virtualization supervisory layers；2) monitoring process is dispatched in virtualization supervisory layers, and when specific process is scheduled, client computer is actively absorbed in virtualization supervisory layers, ad-hoc location injecting codes of the virtualization supervisory layers in client internal memory；3) after Client Restore is performed, perform the code of injection to complete the backup of process status；4) active for realizing client computer by the code of injection is absorbed in, and changes the controlling stream of client computer current process according to the process being absorbed in；5) position of code injection is recovered when needing recovering process context, while waking up the process status of backup.The present invention is backed up with process entity form when process context state needs to be saved in a client, is needed can directly to run the process during recovery after and is resumed work without extra internal memory.
Description
Technical field
The invention belongs to technical field of virtualization, it is related to process status in the real-time preservation client computer based on virtualization technology
Method, more particularly to a kind of being not required to based on virtualization technology change client computer kernel, transparent to client computer process shape
State store method.
Background technology
The topmost effect of virtualization technology is the service efficiency for solving physical resource, and monitor of virtual machine substitutes operation system
Level where system kernel, is carried out abstract to bottom physical resource.Hardware is owned by the distribution and management of virtualization layer by upper strata
Virtual machine is shared, and the virtual machine on upper strata is created, dispatched and distributed resource by virtualization layer.It is virtual on virtualization layer
Machine is mutually isolated, it is necessary to need to be controlled and managed by virtualization layer when mutually accessing or communicate.Virtualization technology is from birth
So far, a variety of correlation techniques and application have been derived.Virtualization technology is due to characteristics, quilt such as its own isolation, access and interventions
It is widely used in safety protection field.These features enable the security tool effective monitoring virtual machine based on virtualization architecture
Internal state, while resisting the attack of monitored internal system.On the whole, rogue program is monitored using virtualization technology and is
System is placed in independent virtualized environment, is responsible for monitoring to virtual machine, can accomplish effectively to resist rogue program hide with
Attack meanses, and easily scalable and maintenance.
For the client computer on virtual platform, the backup of virtual machine state is an important need, especially when
When virtual machine internal or residing platform meet with attack, answer stateful, it is necessary under attack to ensure that virtual machine is quickly recovered to
Some or multiple moment carry out state backup, be rapidly achieved normal operating condition so as to after.To virtual machine state
There are a variety of objects, such as virtual disk, process context, CPU in backup.Common virtual machine backup passes through snapshot, clone etc.
Mode is carried out, and the storage after backup mainly copies to target data in storage device.It is generally necessary to suspend the fortune of virtual machine
OK, backed up internal memory as file, or Memory Backup is realized by the method for dynamic migration.
Backing up in realtime for virtual machine is to solve backup checkpoint and the method for virtual machine current operating conditions inconsistence problems
One of, to accomplish to back up in realtime, it is necessary to any change of virtual-machine data is real-time transmitted on backup server, but data
Transmission be constantly present delay, this is the subject matter backed up in realtime.Cully et al. propose Remus (Cully B,
Lefebvre G,Meyer D,et al.Remus:high availability via asynchronous virtual
machine replication[C]//Usenix Symposium on Networked Systems Design and
Implementation.USENIX Association,2008:It is 161-174.) that the virtual machine based on Xen is backed up in realtime solution
Scheme, it carries out complete copy to virtual machine in file-level, replicates every time just for labeled as dirty page progress, this guarantee
The high efficiency of backup, is consistent virtual machine state and backup file.Back up checkpoint and check virtual by high-frequency
Machine state is determined, then carries out incremental backup to CPU, internal memory and hard disk when capturing the change of execution state.
Parallax(Warfield A,Ross R,Fraser K,et al.Parallax:Managing Storage for a
Million Machines[C]//Hotos'05:,Workshop on Hot Topics in Operating Systems,
June 12-15,2005,Santa Fe,New Mexico,Usa.DBLP,2005:It it is 4-4.) one based on the how virtual of Xen
Machine management system, it is mainly used in solving the virtual machine storage problem of million number of levels, and its Important Thought is all storages
Resource carries out unified addressing, and sets up virtual disk mirror image VDI, each VDI one root node of correspondence, its sub- section for virtual machine
Point is the space for its distribution.Snapshot is pointed to current VDI root nodes when making snapshot, and its all child node is set to
It is read-only.
The backup that mode based on log recording is carried out be mainly used in system it is under attack after playback, it is representative
Achievement in research has Revirt systems (Dunlap G W, King S T, Cinar S, et al.ReVirt:Enabling
intrusion analysis through virtual-machine logging and replay[J].ACM SIGOPS
Operating Systems Review,2002,36(SI):211-224.), it carries out log recording, note on VMM to system
Copyist's part includes asynchronous interrupt, outside input etc., and in terms of being used in intrusion detection, it can be usurped by rogue program in system
In the case of changing, implementation procedure under attack is reset, information is provided for intrusion detection analysis.Revirt needs to grasp client computer
Modified as system kernel, it is opaque to client computer.
Above achievement in research is that process aspect is based on the research of migrating technology on virtual machine-level other backup.Migration
It is, in monitored virtual machine node collapse, to reconfigure and hand on another node and run, migrate usually with process
For unit.The target of migration is to preserve running state of a process, it is can continue to operation in another node, this point and process
There is general character in backup.Process migration need preserve state include the process address space, register context, communications status and
Other information related to the management of process of kernel.MOSIX(Barak A,La'Adan O.The MOSIX
multicomputer operating system for high performance cluster computing[J]
.Future Generation Computer Systems,1998,13(4-5):361-372.) it is by changing operating system
The process migration system that kernel is realized, PRM (Song G, Thomas S, Amato N M.A general framework for
PRM motion planning[C]//IEEE International Conference on Robotics and
Automation,2003.Proceedings.ICRA.IEEE,2003:4445-4450vol.3.) realized based on client layer.This
A little methods belong to full copy type, and recovering process has larger retardance on the target node.
In the above scenario, the research in units of process to be based on migrating, and back up using whole virtual machine state as
It is main.If necessary to which for specific process is backed up and directly backs up whole virtual machine, then larger space expense can be brought.And
With the presence of the certain retardance of process migration technology, be short of at the consistent aspect of process status.
The content of the invention
In view of the above-mentioned problems, the present invention provides process status in a kind of real-time preservation client computer based on virtualization technology
Method.When process context state needs to be saved in some particular moment client computer, it will be carried out with process entity form standby
Part, and when certain moment needs recovering process context after, can directly run the process extensive without extra internal memory
Return to work work.
The technical solution adopted by the present invention is as follows：
A kind of method of process status in real-time preservation client computer based on virtualization technology, its step includes：
1) PID (Process Identification, Process identifier) of specific process in client computer is notified to void
Planization supervisory layers (Virtual Machine Monitor, VMM)；
2) monitoring process is dispatched in VMM, and when specific process is scheduled, client computer is actively absorbed in VMM, and VMM is in client computer
Ad-hoc location injecting codes in internal memory；
3) after Client Restore is performed, perform the code of injection to complete the backup of process status；
4) active for realizing client computer by the code of injection is absorbed in, and client computer current process is changed according to the process being absorbed in
Controlling stream；
5) position of code injection is recovered when needing recovering process context, while waking up the process shape of backup
State.
4 parts are classified into below the particular content of above step is described in detail.
1. after VMM receives the PID of specific process in client computer, the task_struct structures of process are obtained according to PID,
And then obtain page directory tables address, the i.e. value of cr3 registers.The monitoring to cr3 register writes is then turned on, when cr3 is posted
When storage is written into the page directory tables address of specified process, then start to ad-hoc location injecting codes in client internal memory.
2. the function that the code of injection is realized, which includes injection particular system, calls and realizes that the active of client computer is absorbed in.
Linux system is called to traditional system using 0x80 implement of interruption function, and although redaction has sysenter/syscall and refer to
The rapid system call instruction of order, but still compatibility is instructed to int80, therefore this method selection int80 instruction execution systems are adjusted
Injection.The reason for client computer is absorbed in has a variety of, and such as client computer performs privileged instruction, and client computer accesses physical resource etc., such as
Fruit selects the conventional reason that is absorbed in repeatedly invalid situation about being absorbed in then easily occur.Client computer is absorbed in reason and can controlled by VMM
System a, part is absorbed in reason and is defaulted as being not turned on being absorbed in VMM.What this method selection VMM acquiescences were not turned on is absorbed in reason, and
Client computer itself will not frequently occur this mode for being absorbed in reason and realize that active is absorbed in, and select int3 instructions to realize here.
3. the position selection of injecting codes will perform the generation of injection in user mode codes section after Client Restore right of execution
Code, and complete the backup of process status.The backup process of generation needs to be specially treated, because backup process is also to be adjusted
The process of degree, to prevent backup process from losing context state, it is necessary to change the scheduling to backup process, this method is used and is absorbed in
The mode that injection particular system is called after VMM realizes the change to the process scheduling situation.
4. the context of process needs to be resumed, then change the dispatch situation of backup process again, allow the process energy quilt
Normal consistency.The position of injecting codes in internal memory is recovered simultaneously.
The beneficial effects of the invention are as follows：
When process context state needs to be saved in client computer, client computer need not assist VMM to do state and preserve work
Make, i.e., running state of process is unaffected in client computer；When certain moment needs recovering process context after, Neng Gouzhi
Connect operation the process resume work without extra internal memory.In whole process, client computer will be unaware that controlling stream
Change, and client computer kernel code need not be modified and be installed kernel module.
Brief description of the drawings
Fig. 1 is that process status of the present invention preserves process chart.
Fig. 2 is that process status of the present invention recovers process chart.
Embodiment
The present invention is explained in further detail with reference to embodiment and accompanying drawing.It is understood that described
Embodiment is only a part of embodiment of the invention, rather than whole embodiments.Based on the embodiment in the present invention, this area
The every other embodiment that technical staff is obtained under the premise of creative work is not made, belongs to the model that the present invention is protected
Enclose.
Embodiment one：
It is that one fork system of process injection is called, when Client Restore is held to carry out state backup to specific process
This system will be performed during row call and automatically create out a subprocess, here using this subprocess as backup process, and father enters
Cheng Zhengchang is performed.Parent process is called after the subprocess that fork is created that, in addition it is also necessary to call wait functions to wait subprocess state to become
Change, wait functions can block parent process, when subprocess end of run, wait functions are returned, parent process receives subprocess
It is subprocess Resource recovery after the signal of end of run.If parent process does not call wait functions to wait subprocess running status
Change, then resource will not be recovered and turn into zombie process after subprocess end of run.If however, parent process is prior to subprocess
End of run, then a process, i.e. init processes can turn into the parent process of subprocess, and be its time in subprocess end of run
Receive resource.The system calls carry out process backup using fork systems, and the backup that follow-up parent process is bound to prior to its establishment is entered
Journey terminates operation, therefore backup process will not be changed into zombie process, and its resource can be reclaimed correctly.
To change, client computer is ensuing to perform flow, it is necessary to change 4 bytes of process code section.Client is obtained first
Eip register values when machine is absorbed in, i.e., the linear address for the next instruction that will be performed.Then four started herein are changed
Byte of memory, is changed to int3, int80, int3, the machine code of wherein int3 instructions is CC, and int80 machine code is
CD80, three instructions just take four bytes, and amended nybble content is CC CD 80CC.Wherein, int80 instructions are used
Called to produce fork systems, the active that int3 instructions are used for producing client computer to VMM is absorbed in, and needs exist for the active of client computer
It is that the code changed needs to be resumed to be absorbed in reason.Also need to recover the content of this section of code script due to after, because
It is noted that preserving the original content in the position so that follow-up is resumed work before this modification.
Why to inject two int3 and interrupt makes client computer active generation be absorbed in twice, is because modification is that internal memory is complete
The code of office, other unrelated processes may also share this one page and perform the code that this section is changed, if allowing them all to go to hold
This section of code of row can increase complexity.At the time of the time point being resumed at code revision is that backup process is released, if only
One int3 of injection is interrupted, and before this, as long as unrelated process is gone to here, all can first go to perform int80 then due to
Int3 and actively be absorbed in, if at this moment again go recover the unrelated process context, the value for facing eax registers has been lost
The problem of losing and can not recover.Therefore int80 execution before, more injection one int3 make its system call execution before just
Actively it is absorbed in, so VMM can be allowed to be judged according to the situation of being absorbed in and make correct processing.The processing being absorbed in after VMM is shown in
Fig. 1, as shown in the drawing, if be absorbed in from first int3, no matter the process is unrelated process or controlled process, is required to mould
Intend the original code implementation effect in the position changed；If be absorbed in from second int3, judge the process for controlled process
Or the subprocess that backup is produced, if controlled process then records function call and returns to the subprocess PID produced, if subprocess
Then specifically handled according to whether discharging the process.
Embodiment two：
To realize that process status is backed up, the process being created that should be suspended in this state, without the execution that should be scheduled.Place
Reason process is as shown in Figure 2.To make process not be scheduled, called using injection sched_yield systems allow the process to sleep here.
The same fork of method for implanting, that is, it is sched_yield system call numbers to change eax registers, and changing eip registers points to it
Int80 IAes.If detecting backup process to be scheduled, sched_yield is injected, until needing recovering process.
If VMM receives process status and recovers signal, when generation backup process is absorbed in, it is necessary to recover its fortune
OK, injection sched_yield systems are now needed not continue to call, but the code content of modification is recovered.However,
Be now in dormant process it is more than this, it is also possible to there is the backup process that other needs are resumed.If
Judge to need to be resumed when one backup process is absorbed in recover original code content with regard to horse back, then other backup process are gone to
, will be because mistake occurs in context and can not continue to run with when herein.Therefore, when first backup process is absorbed in, it is impossible to
Recover original code content at once, but original code implementation effect is first simulated for it, by corresponding registers, it is specific in
Deposit and change into the content after code is performed, called while not reinjecting sched_yield systems.This backup process subsequently will not
It can be absorbed in due to second int3.When the process of recovery in need be all waken up, i.e., last backup process exists
When second int3 is absorbed in, then recover original code.Code recover after, will not any further process because int3 occur
It is absorbed in.
Claims (8)
1. a kind of method of process status in real-time preservation client computer based on virtualization technology, its step includes：
1) PID of specific process in client computer is notified to give virtualization supervisory layers；
2) monitoring process is dispatched in virtualization supervisory layers, and when specific process is scheduled, client computer is actively absorbed in virtualization monitoring
Layer, ad-hoc location injecting codes of the virtualization supervisory layers in client internal memory；
3) after Client Restore is performed, perform the code of injection to complete the backup of process status；
4) active for realizing client computer by the code of injection is absorbed in, and changes the control of client computer current process according to the process being absorbed in
System stream；
5) position of code injection is recovered when needing recovering process context, while waking up the process status of backup.
2. the method as described in claim 1, it is characterised in that：Virtualization supervisory layers receive the PID of specific process in client computer
Afterwards, the task_struct structures of process are obtained according to PID, and then obtain page directory tables address, the i.e. value of cr3 registers；Then
The monitoring to cr3 register writes is opened, when cr3 registers are written into the page directory tables address of specified process, is then started
To ad-hoc location injecting codes in client internal memory.
3. the method as described in claim 1, it is characterised in that：The injection and client that system calls are realized by the code of injection
The active of machine is absorbed in.
4. method as claimed in claim 3, it is characterised in that：The injection called by int80 instruction execution systems, passes through
Int3 instructions realize that the active of client computer is absorbed in.
5. method as claimed in claim 4, it is characterised in that the method for injecting codes is in a client：Visitor is obtained first
Eip register values when family machine is absorbed in, i.e., the linear address for the next instruction that will be performed；Then four started herein are changed
Individual byte of memory, is changed to int3, int80, int3, and wherein int80 instructions are called for producing fork systems, int3 instructions
It is absorbed in for producing client computer to the active for virtualizing supervisory layers.
6. the method as described in claim 1, it is characterised in that：Called using injection special system after virtualization supervisory layers is absorbed in
Mode realize the change to backup process dispatch situation, with prevent backup process lose context state；Above and below process
When text needs to be resumed, change the dispatch situation of backup process again, so that the process can be by normal consistency.
7. method as claimed in claim 6, it is characterised in that：Make entering for backup by injecting sched_yield systems and calling
Journey is slept, and prevents it from the execution that is scheduled, until needing recovering process.
8. method as claimed in claim 7, it is characterised in that：If virtualization supervisory layers, which receive process status, recovers letter
Number, then when generation backup process is absorbed in, recover its operation；If being in dormant backup process more than one, when
Be first that it simulates original code implementation effect when first backup process is absorbed in, when the process of recovery in need all called out
When waking up, then recover original code.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710233176.5A CN107239320A (en) | 2017-04-11 | 2017-04-11 | The method of process status in real-time preservation client computer based on virtualization technology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710233176.5A CN107239320A (en) | 2017-04-11 | 2017-04-11 | The method of process status in real-time preservation client computer based on virtualization technology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107239320A true CN107239320A (en) | 2017-10-10 |
Family
ID=59983337
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710233176.5A Pending CN107239320A (en) | 2017-04-11 | 2017-04-11 | The method of process status in real-time preservation client computer based on virtualization technology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107239320A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020147544A1 (en) * | 2019-01-18 | 2020-07-23 | 华为技术有限公司 | Method and device for resuming execution of application, and computer |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040268347A1 (en) * | 2003-06-26 | 2004-12-30 | Knauerhase Robert C. | Virtual machine management using processor state information |
| CN101405712A (en) * | 2006-03-21 | 2009-04-08 | 英特尔公司 | Framework for domain-specific run-time environment acceleration using virtualization technology |
| US20090276772A1 (en) * | 2008-04-30 | 2009-11-05 | Garrett Michael R | Virtualization platform configured with virtual connect control |
| CN101968746A (en) * | 2010-09-02 | 2011-02-09 | 北京航空航天大学 | Method for implementing organizational architecture mode of kernel-based virtual machine (KVM) |
| CN103955438A (en) * | 2014-05-21 | 2014-07-30 | 南京大学 | Process memory protecting method based on auxiliary virtualization technology for hardware |
| CN104102528A (en) * | 2014-07-14 | 2014-10-15 | 南京理工大学 | Quick performance restoration method of virtual machine monitor |
-
2017
- 2017-04-11 CN CN201710233176.5A patent/CN107239320A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040268347A1 (en) * | 2003-06-26 | 2004-12-30 | Knauerhase Robert C. | Virtual machine management using processor state information |
| CN101405712A (en) * | 2006-03-21 | 2009-04-08 | 英特尔公司 | Framework for domain-specific run-time environment acceleration using virtualization technology |
| US20090276772A1 (en) * | 2008-04-30 | 2009-11-05 | Garrett Michael R | Virtualization platform configured with virtual connect control |
| CN101968746A (en) * | 2010-09-02 | 2011-02-09 | 北京航空航天大学 | Method for implementing organizational architecture mode of kernel-based virtual machine (KVM) |
| CN103955438A (en) * | 2014-05-21 | 2014-07-30 | 南京大学 | Process memory protecting method based on auxiliary virtualization technology for hardware |
| CN104102528A (en) * | 2014-07-14 | 2014-10-15 | 南京理工大学 | Quick performance restoration method of virtual machine monitor |
Non-Patent Citations (1)
| Title |
|---|
| 李津津、贾晓启、杜海超1、王利朋: "基于虚拟化技术的有效提高***可用性的方法", 《计算机应用》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020147544A1 (en) * | 2019-01-18 | 2020-07-23 | 华为技术有限公司 | Method and device for resuming execution of application, and computer |
| CN111459623A (en) * | 2019-01-18 | 2020-07-28 | 华为技术有限公司 | Method and device for restoring running of application program and computer |
| CN111459623B (en) * | 2019-01-18 | 2024-04-12 | 华为技术有限公司 | Method, device and computer for restoring running of application program |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10884870B2 (en) | Method and system for implementing consistency groups with virtual machines | |
| US7313793B2 (en) | Method for forking or migrating a virtual machine | |
| US9417965B2 (en) | Low overhead fault tolerance through hybrid checkpointing and replay | |
| EP2053509B1 (en) | System for and method of migrating one or more virtual machines | |
| US8495316B2 (en) | Efficient management of archival images of virtual machines having incremental snapshots | |
| US7627728B1 (en) | System and method for efficient generation of application snapshots | |
| CN104407938B (en) | A kind of a variety of granularity restoration methods after virtual machine image level backup | |
| US9852198B1 (en) | Method and system for fast generation of file system snapshot bitmap in virtual environment | |
| CN101377750B (en) | System and method for cluster fault toleration | |
| US8121977B2 (en) | Ensuring data persistence and consistency in enterprise storage backup systems | |
| US7669020B1 (en) | Host-based backup for virtual machines | |
| CN103793288B (en) | A kind of software watchdog system and method | |
| CN111708488B (en) | Distributed memory disk-based Ceph performance optimization method and device | |
| US20230161674A1 (en) | Live Migrating Virtual Machines to a Target Host Upon Fatal Memory Errors | |
| CN107239320A (en) | The method of process status in real-time preservation client computer based on virtualization technology | |
| CN111459607A (en) | Virtual server cluster building method, system and medium based on cloud desktop virtualization | |
| US20230121338A1 (en) | Virtual Machines Recoverable From Uncorrectable Memory Errors | |
| Rosenblum et al. | Implementing efficient fault containment for multiprocessors: confining faults in a shared-memory multiprocessor environment | |
| Korotkevitch | SQL Server Advanced Troubleshooting and Performance Tuning | |
| Yao et al. | Using Block Filter Device for Acceleration of Storage Synchronization on Fault-tolerant Virtual Machines | |
| US7761737B2 (en) | Apparatus, method, and computer product for executing program | |
| Xu et al. | Steady rollback and recovery policy based on integrity measurement |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171010 |
|
| RJ01 | Rejection of invention patent application after publication |