IBM Storwize V7000 Introduction(1)
先看一段來(lái)自redbook的Strowize V7000的定義:An IBM Storwize V7000 system is a clustered, scalable, and midrange storage system, and an external virtualization device.
V7000最大的賣點(diǎn)是Virtualization。所謂的存儲(chǔ)虛擬化是將不同廠商的存儲(chǔ)設(shè)備放到一個(gè)Pool里以便統(tǒng)一管理,例如下圖Storwize V7000(SVC)后面可以接IBM的DS8000,XIV,NetApp,EMC等其他廠商的設(shè)備。將這些設(shè)備的所有存儲(chǔ)空間放到一個(gè)大的Pool里,跨設(shè)備之間做copy service,Data Migration。對(duì)后端存儲(chǔ)做操作時(shí)對(duì)前端主機(jī)是透明的,因?yàn)榍岸薶ost只看到一個(gè)Storwize V7000(or SVC)。
?
說(shuō)到虛擬化,IBM除了Storwize V7000,還有SVC(SAN Volume controller)。兩者運(yùn)行的code都是一樣的,只不過(guò)一個(gè)是作為Server(SVC),一個(gè)是作為Storage(Storwize V7K)。下圖是一套Storwize V7000系統(tǒng)的組成部分,類似于DS3K/4K/5K,它也分Control Enclosure和Expansion Enclosure。
?
再看看Storwize V7000的硬件,Front view和Font view,硬件外觀上看上去和DS3512很像,只不過(guò)Storwize V7000中叫做Node Canisters,而DS3512叫做Controller。每個(gè)Storwize V7000有一個(gè)或兩個(gè)(optional)control enclosures,每個(gè)control enclosure都包括兩個(gè)node canisters,磁盤及兩個(gè)PSU。
?
重要的術(shù)語(yǔ):
Control enclosure | A hardware unit that includes the chassis with a midplane for connection of node canisters, drives and power supplies with batteries. |
Node canister 【類似于DS5K的Controller】 | A hardware unit that includes the node electronics, fabric and service interfaces, serial-attached SCSI (SAS) expansion ports and direct connections to internal drives in the enclosure. |
Expansion enclosure | A hardware unit that includes the chassis with a midplane for connection of expansion canisters, drives, and power supplies without batteries. |
Expansion canister 【類似于DS5K擴(kuò)展柜的ESM】 | A hardware unit that includes the electronics to provide serial-attached SCSI (SAS) connections to the internal drives in the enclosure and SAS expansion ports for p_w_upload of additional expansion enclosures. |
Cluster | Two node canisters in a control enclosure. |
Managed Disk (MDisk) | A SCSI logical unit (aka LUN) built from an internal or external RAID array. |
Storage pool | A collection of MDisks providing real capacity for volumes. SVC: Managed Disk Group (MDG) |
Volume | What the host operating system sees as a SCSI disk drive. SVC: Virtual Disk (VDisk) |
Cluster ? ?
Cluster由2~8個(gè)節(jié)點(diǎn)組成,所有的配置,監(jiān)控和服務(wù)進(jìn)程是在Cluster level執(zhí)行,configuration 是copy在Cluster里的所有node 上。Cluster只分配一個(gè)IP而不是每個(gè)node一個(gè)IP。其中一個(gè)Node會(huì)被選出成為“configuration node canister”,它是唯一一個(gè)來(lái)activate Cluster IP的node。如果這個(gè)node fail了,會(huì)重新選出新的“configuration node”,新的“configuration node”會(huì)接管這個(gè)IP。
I/O Groups:兩個(gè)節(jié)點(diǎn)構(gòu)成的類似AB控制器的一對(duì), VDisk在同一個(gè)IOgroup的兩個(gè)節(jié)點(diǎn)之間fail over, fail back
正常情況下,某個(gè)Volume的IO都是由I/O group里的同一個(gè)Node處理的。和DS3K/4K/5K一樣,一個(gè)Volume也只有一個(gè)Preferred Owner(A控orB控,Node1orNode2),兩個(gè)Node是以failover mode模式工作,即一個(gè)Node down掉時(shí),另一個(gè)Node可以繼續(xù)工作,而對(duì)主機(jī)沒(méi)有impact。
?
一個(gè)Node就是一個(gè)SVC(SAN Volume controller),是個(gè)運(yùn)行Linux的X Series服務(wù)器,提供虛擬化,copy service。兩個(gè)Node組成一個(gè)Cluster。一個(gè)Cluster可以有1-4個(gè)Node paris。每個(gè)pair都是一個(gè)IO Group。IO group在配置Cluster的時(shí)候定義。每個(gè)Node只能在一個(gè)IO group里。
Managed Disk (MDisk)
Host端不會(huì)看到Managed disk,相反,他們看到的是Logical disks,也被稱為Virtual disks。Managed disks再組成managed disk groups。組成Virtual disks的Managed disk必須來(lái)自同一個(gè)Managed disk group。每個(gè)Managed disk被分成多個(gè)Extents(默認(rèn)16MB,16MB~8192MB),從0開(kāi)始,一直到Managed disk的結(jié)束。建議一個(gè)system里所有Storage Pool都使用一個(gè)Extent,這是在兩個(gè)Storage Pool之間做volume Migration的前提,當(dāng)然如果extent不一樣,可以使用volume Mirroring。
關(guān)于Mdisk,Storage Pool,Vdisk之間的關(guān)系,下圖闡釋的非常清楚。
Storage Pool
多個(gè)Mdisks(最多為128個(gè))組成一個(gè)Storage Pool。
Quorum disk,Configuration Information是存在QuorumDisk上的,system里的所有node都可訪問(wèn),Quorumdisk可以是一個(gè)Mdisk或者是Internal Drive,也是仲裁盤。一個(gè)Cluster由3個(gè)quorum disk,但只有一個(gè)是Active的。在V7K,Internal drive可以被認(rèn)為是Quorum candidates。如果一個(gè)環(huán)境里有多個(gè)存儲(chǔ),那為了避免單個(gè)Storage壞掉而丟失Quorum disk,應(yīng)該把Quorum disk分散在多個(gè)存儲(chǔ)上。
Quorum disk是來(lái)決定當(dāng)Cluster中的一個(gè)node offline 時(shí),哪個(gè)node繼續(xù)工作。In this tie-break situation, the first group of nodes that access the quorum disk marks their ownership of the quorum disk and as a result continues to operate as the cluster. If the other group of nodes cannot access the quorum disk or finds it owned by another group of nodes it stops operating as the cluster and does not handle I/O requests.
?
看下一個(gè)write request是怎么在V7000中完成的:
創(chuàng)建Vdisk時(shí)就已經(jīng)define了preferred node,在上圖中即node1,正常情況下所有對(duì)V1這個(gè)Vdisk的讀寫操作都是由node1完成的。
通過(guò)Node1 對(duì)Vdisk V1進(jìn)行寫操作請(qǐng)求(1),write data會(huì)copy一份到node2里,然后才會(huì)給主機(jī)返回一個(gè)寫完成操作(2)。之后,node1會(huì)將cache數(shù)據(jù)destage到disk
總結(jié):寫操作時(shí),處理IO的Node會(huì)將數(shù)據(jù)copy一份到IO group里的另一個(gè)Node,然后才會(huì)給主機(jī)返回一個(gè)write completion。讀數(shù)據(jù)時(shí),和其他storage一樣,preferred node會(huì)先檢查cache中有沒(méi)有data。如果沒(méi)有就從磁盤中讀。如果IO group里的一個(gè)node fail了,另一個(gè)node會(huì)立刻接管過(guò)來(lái),由于write data是在兩個(gè)node之間mirror的,所以這種node failure不會(huì)造成data loss。Surviving node會(huì)將cache里的數(shù)據(jù)destage到disk并進(jìn)入write-through mode,即所有write data直接寫到disk而不經(jīng)過(guò)cache。
轉(zhuǎn)載于:https://blog.51cto.com/taotao1240/1599615
總結(jié)
以上是生活随笔為你收集整理的IBM Storwize V7000 Introduction(1)的全部?jī)?nèi)容,希望文章能夠幫你解決所遇到的問(wèn)題。
- 上一篇: js中document.write()使
- 下一篇: 比亚迪“大爆发”!半年净利润超105亿,