Summary of the invention
The present invention provides a kind of upgrade methods and system based on classification storage, cannot achieve greatly with solving the prior artCapacity, continuity, reliability and multifarious defect.
The present invention provides a kind of upgrade methods based on classification storage, comprising the following steps:
Using living document system ZFS as underlying file systems, cluster resource manager pacemaker and heartbeat are configuredCorosync redundancy structure is as basic storage system；
In the environment of the basic storage system, the network attached storage NAS system based on Network File System NFS is usedSystem is used as hierarchical stor.
Optionally, described using living document system ZFS as underlying file systems, configure cluster resource managerPacemaker and heartbeat corosync redundancy structure are used as before basic storage system, further includes:
System upgrade environment is set.
Optionally, the setting system upgrade environment, specifically includes:
Network ip address is arranged in the network at configuration server end, realizes that safety is outer between Lustre node and NFS nodeSSH is without cryptographic acess for shell agreement；The firewall of all nodes is closed, SELinux method is disabled；By shared storage mapping to accordinglyLustre node on, installation Lustre software package, ZFS software package, pacemaker software package, corosync software package and NFSSoftware package.
Optionally, described in the environment of the basic storage system, it is attached using the network based on Network File System NFSAfter belonging to storage NAS system as hierarchical stor, further includes:
Based on the hierarchical stor, compatible several operation systems are as NFS client.
Optionally, described to be based on the hierarchical stor, compatible several operation systems as NFS client, specifically:
In the environment of the NAS high capacity based on NFS is classified space, compatible windows, linux and mac operating system is madeFor NFS client.
The present invention also provides a kind of upgrade-systems based on classification storage, comprising:
First configuration module, for configuring cluster resource manager using living document system ZFS as underlying file systemsPacemaker and heartbeat corosync redundancy structure are as basic storage system；
Second configuration module is used in the environment of the basic storage system, using based on Network File System NFS'sNetwork attached storage NAS system is as hierarchical stor.
Optionally, the system, further includes:
Setup module, for system upgrade environment to be arranged.
Optionally, network ip address is arranged specifically for the network at configuration server end in the setup module,Realize safety shell protocol SSH without cryptographic acess between Lustre node and NFS node；The firewall of all nodes is closed, is prohibitedWith SELinux method；Will in shared storage mapping to corresponding Lustre node, install Lustre software package, ZFS software package,Pacemaker software package, corosync software package and NFS software package.
Optionally, the system, further includes:
Compatible modules, for being based on the hierarchical stor, compatible several operation systems are as NFS client.
Optionally, the compatible modules, specifically for being compatible in the environment of the NAS high capacity based on NFS is classified spaceWindows, linux and mac operating system are as NFS client.
The present invention builds Lustre using ZFS as underlying file systems, can guarantee the high bandwidth performance of storage system, makeTwo-shipper High Availabitity is done with pacemaker and corosync, can guarantee the fast failure recovery ability of system and continuing for businessProperty uses NAS storage system as the storage system of high capacity, can satisfy the storage of Large Volume Data, and needs whenIt waits and adjustable uses high band broad base storage system.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, completeSite preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based onEmbodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every otherEmbodiment shall fall within the protection scope of the present invention.
It should be noted that each feature in the embodiment of the present invention and embodiment can be tied mutually if do not conflictedIt closes, it is within the scope of the present invention.In addition, though logical order is shown in flow charts, but in certain situationsUnder, it can be with the steps shown or described are performed in an order that is different from the one herein.
The embodiment of the invention provides a kind of upgrade methods based on classification storage, as shown in Figure 1, comprising the following steps:
Step 101, system upgrade environment is set.
Specifically, network ip address can be set, in Lustre node and NFS with the network at configuration server endSSH (Secure Shell Protocol, safety are realized between (Network File System, Network File System) nodeShell protocol) without cryptographic acess, guarantee the network connection between server end；The firewall of all nodes is closed, is disabledSELinux method；By in shared storage mapping to corresponding Lustre node, Lustre software package, ZFS (Zettabyte are installedFile System, living document system) software package, pacemaker (cluster resource manager) software package, corosync (heartbeatLine) software package and NFS software package.
Step 102, using ZFS as underlying file systems, based on configuration pacemaker and corosync redundancy structureStorage system.
Specifically, parameter backfstype=zfs can be added, and separately do two-shipper High Availabitity and set when formatting hard diskFailnode is set, such as:
[root@node02~] #mkfs.lustre -- reformat -- mgs -- mdt -- backfstype=zfs-- fsname=qq--failnode=node01@tcp0 -- index=0 lustre-mgs/mgt raidz/dev/sd a,b,c}
[root@node01~] #mkfs.lustre -- reformat -- ost -- index=0 -- backfstype=zfs -- mgsnode=node02@tcp0--failnode=node02@tcp0 -- fsname=qq lustre-ost0/ost0raidz2/dev/sd{a,b,c,d,e}
Start Lustre service, carry Lustre client:
[root@cient1~] #mount-t lustre node02@tcp0:node01@tcp0:/qq/mnt/qq
Corosync configuration file is distributed to Lustre service end node, starts pacemaker and corosync service,It checks crm_mon, obtains following result:
Last updated:Thu Nov 6 17:13:49 2014
Last change:Thu Nov 6 15:39:36 2014 via crmd on node02
Stack:classic openais(with plugin)
Current DC:node02-partition with quorum
Version:1.1.10-14.el6_5.3-368c726
2 Nodes configured,2 expected votes
4 Resources configured
Online:[node01node02]
mgt (ocf::heartbeat:Lustre): Started node01
ost0 (ocf::heartbeat:Lustre): Started node02
Clone Set:clonePing[resPing]
Started:[node01node02]
Step 103, in the environment of basic storage system, NAS (the Network Attached based on NFS is usedStorage, network attached storage) system is as hierarchical stor.
Specifically, export file can be configured in NFS server-side:
/ mnt/nfstest* (fsid=0, rw, async, insecure, no_root_squash, no_subtree_check)
Start NFS service, carry out carry in client:
mount-t nfs node03:/mnt/nfstest/mnt/nfs
Coordinator is activated on lustre mds node
[root@m2~] #lctl set_param mdt.testfs-MDT0000.hsm_control=enabled
Mdt.testfs-MDT0000.hsm_control=enabled
[root@m2~] #lctl get_param mdt.testfs-MDT0000.hsm_control
Mdt.testfs-MDT0000.hsm_control=enabled
Start agent
[root@node01~] #lhsmtool_posix -- daemon -- hsm-root/nfs -- archive=1/testfs
Lhsmtool_posix [32510]: action=0src=(null) dst=(null) mount_point=/testfs
hsmtool_posix[32511]:waiting for message from kernel
Activation release is carried out to file using hsm order and checks state
[root@client02hsmfs]#lfs hsm_archive–archive 1 file
[root@client02hsmfs]#lfs hsm_state file
[root@client02hsmfs]#lfs hsm_release file
Step 104, it is based on hierarchical stor, compatible several operation systems are as NFS client.
Specifically, can in the environment of the NAS high capacity based on NFS is classified space, compatible windows, linux andMac operating system checks user to file by NFS client and data is transmitted, satisfaction as NFS clientThe demand of different operating system operation.
In the present embodiment, carry can be carried out to linux system in the following ways:
Start NFS service, establish mount directory, carry NFS client:
#mount–t nfs 100.1.1.10:/mnt/nfs/mnt/nfs
Carry can also be carried out to Windows system in the following ways:
Control panel program open or close windows function under, NFS client and tool are installed；It is transported in cmdRow:
C:\Users\pc>showmount-e 100.1.1.10
List is exported in 100.1.1.10:
/home *
/opt *
/mnt/nfs *
C:\Users\pc>mount\\100.1.1.10\mnt\nfs Z:
Z: be successfully connected to 100.1.1.10 mnt nfs
Order is successfully completed.
It should be noted that need to be applicable in some softwares in the windows system of other certain versions and be hung to runIt carries, such as SFU software.
Further, it is also possible to carry out carry to Mac system in the following ways:
Nfs server-side can be connected in terminal detection:
showmount-e 100.1.1.10
exports list on100.1.1.10:
/home *
/opt *
/mnt/nfs
Execute carry order:
mount–t nfs 100.1.1.10:/mnt/nfs/private/nfs
If display reports an error: operation not permitted
Then modify carry order:
mount–t nfs–o resvport,rw 100.1.1.10:/mnt/nfs/private/nfs
The embodiment of the present invention builds Lustre using ZFS as underlying file systems, can guarantee the high bandwidth of storage systemPerformance does two-shipper High Availabitity using pacemaker and corosync, can guarantee the fast failure recovery ability and business of systemDuration, use NAS storage system as the storage system of high capacity, can satisfy the storage of Large Volume Data, and needingIt is adjustable when wanting to use high band broad base storage system.In addition, the second order hierarchical space using NAS system as high capacity, energyEnough compatible windows, linux and mac system client, to meet the storage that user carries out file using different operating systemWith the demand of transmission, the diversity of user is improved.
Based on the above-mentioned upgrade method based on classification storage, the embodiment of the invention also provides a kind of based on classification storageUpgrade-system, as shown in Figure 2, comprising:
First configuration module 201, for configuring pacemaker and corosync redundancy using ZFS as underlying file systemsFramework is as basic storage system；
Second configuration module 202 uses the NAS system based on NFS as dividing in the environment of basic storage systemGrade storage system.
Further, above system, further includes:
Setup module 203, for system upgrade environment to be arranged.
Specifically, network ip address is arranged specifically for the network at configuration server end in above-mentioned setup module 203,Realize safety shell protocol SSH without cryptographic acess between Lustre node and NFS node；The firewall of all nodes is closed, is prohibitedWith SELinux method；Will in shared storage mapping to corresponding Lustre node, install Lustre software package, ZFS software package,Pacemaker software package, corosync software package and NFS software package.
Further, above system, further includes:
Compatible modules 204, for being based on the hierarchical stor, compatible several operation systems are as NFS client.
Specifically, above-mentioned compatible modules, specifically for being compatible in the environment of the NAS high capacity based on NFS is classified spaceWindows, linux and mac operating system are as NFS client.
The embodiment of the present invention builds Lustre using ZFS as underlying file systems, can guarantee the high bandwidth of storage systemPerformance does two-shipper High Availabitity using pacemaker and corosync, can guarantee the fast failure recovery ability and business of systemDuration, use NAS storage system as the storage system of high capacity, can satisfy the storage of Large Volume Data, and needingIt is adjustable when wanting to use high band broad base storage system.In addition, the second order hierarchical space using NAS system as high capacity, energyEnough compatible windows, linux and mac system client, to meet the storage that user carries out file using different operating systemWith the demand of transmission, the diversity of user is improved.
Step in method described in conjunction with the examples disclosed in this document can directly use hardware, processor to executeThe combination of software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only memory(ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical fieldIn any other form of storage medium well known to interior.
The above description is merely a specific embodiment, but scope of protection of the present invention is not limited thereto, anyThose familiar with the art in the technical scope disclosed by the present invention, can easily think of the change or the replacement, and should all containLid is within protection scope of the present invention.Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.