Using Red Hat Clustering Suite or SUSE Linux on a vSphere 6.x vSAN Datastore

안녕하세요. vSAN 환경에서 SUSE HA 를 구성할 경우에 대한 문의가 있어서 한번 확인해보았습니다.

다행히 잘 정리된 KB 가 있네요. 

https://kb.vmware.com/s/article/2151774

추가적으로 몰랐던 사실은 vSphere 6.5 에서는 OSR 을 변경하지 않더라도 Thick eager 또는 lazy zeroed 설정할 수 있었다는 것입니다. (아래의 Disk Provisioning 부분)

실제로 어떻게 생성이 되는지 확인해봤습니다. (40GB 디스크로 Thick Eager Zeroed 로 Provisioning)

Disk backing: [con01-vsan01] 0a4a7d5a-28b0-1a8a-7488-b8aeedeaf9b6/TEST.vmdk
DOM Object: 0b4a7d5a-c4ec-4e30-f58c-b8aeedeaf9b6 (v5, owner: esxi01-con.psolab.local, proxy owner: None, policy: spbmProfileId = aa6d5a82-1c88-45da-85d3-3d74b91a5bad, cacheReservation = 0, hostFailuresToTolerate = 1, spbmProfileGenerationNumber = 0, proportionalCapacity = 100, forceProvisioning = 0, spbmProfileName = vSAN Default Storage Policy, CSN = 2, stripeWidth = 1)
RAID_1
Component: 0b4a7d5a-5046-2631-9a1e-b8aeedeaf9b6 (state: ACTIVE (5), host: esxi01-con.psolab.local, md: t10.ATA_____Samsung_SSD_850_PRO_512GB_______________S257NX0H600112Z_____, ssd: t10.ATA_____Samsung_SSD_850_EVO_M.2_120GB___________S24ANX0H605235X_____,
votes: 1, usage: 40.8 GB, proxy component: false)
Component: 0b4a7d5a-0ea2-2731-b01a-b8aeedeaf9b6 (state: ACTIVE (5), host: esxi02-con.psolab.local, md: t10.ATA_____Samsung_SSD_850_PRO_512GB_______________S257NX0H600114B_____, ssd: t10.ATA_____Samsung_SSD_850_EVO_M.2_120GB___________S24ANX0H605242J_____,
votes: 1, usage: 40.8 GB, proxy component: false)
Witness: 0b4a7d5a-b46a-2831-c758-b8aeedeaf9b6 (state: ACTIVE (5), host: vsanwit.psolab.local, md: mpx.vmhba1:C0:T1:L0, ssd: mpx.vmhba1:C0:T2:L0,
votes: 1, usage: 0.0 GB, proxy component: false)

Thin Provisioning 으로 한 경우

Disk backing: [con01-vsan01] b24c7d5a-f4fd-f91a-1811-b8aeedeaf9b6/TEST2.vmdk
DOM Object: b34c7d5a-5eb5-a7ba-f8ba-b8aeedeaf9b6 (v5, owner: esxi01-con.psolab.local, proxy owner: None, policy: spbmProfileName = vSAN Default Storage Policy, cacheReservation = 0, proportionalCapacity = 0, hostFailuresToTolerate = 1, CSN = 1, forceProvisioning = 0, stripeWidth = 1, spbmProfileGenerationNumber = 0, spbmProfileId = aa6d5a82-1c88-45da-85d3-3d74b91a5bad)
RAID_1
Component: b34c7d5a-e892-57bb-5e07-b8aeedeaf9b6 (state: ACTIVE (5), host: esxi01-con.psolab.local, md: t10.ATA_____Samsung_SSD_850_PRO_512GB_______________S257NX0H600112Z_____, ssd: t10.ATA_____Samsung_SSD_850_EVO_M.2_120GB___________S24ANX0H605235X_____,
votes: 1, usage: 0.0 GB, proxy component: false)
Component: b34c7d5a-d6ce-58bb-dfa5-b8aeedeaf9b6 (state: ACTIVE (5), host: esxi02-con.psolab.local, md: t10.ATA_____Samsung_SSD_850_PRO_512GB_______________S257NX0H600114B_____, ssd: t10.ATA_____Samsung_SSD_850_EVO_M.2_120GB___________S24ANX0H605242J_____,
votes: 1, usage: 0.0 GB, proxy component: false)
Witness: b34c7d5a-1072-59bb-27dd-b8aeedeaf9b6 (state: ACTIVE (5), host: vsanwit.psolab.local, md: mpx.vmhba1:C0:T1:L0, ssd: mpx.vmhba1:C0:T2:L0,
votes: 1, usage: 0.0 GB, proxy component: false)

6.5 이전버전에서는 UI 상에서는 불가능했는데, 참고하시면 되겠습니다.

Update – 2018/03/13

혹시나 사용하는 Cluster Solution 이 SCSI3 PR(persistent reservations) 을 Must 로 요구하는 경우에는 vSAN 으로는 불가능합니다. (RDM 만 가능)

답글 남기기

이메일 주소는 공개되지 않습니다. 필수 필드는 *로 표시됩니다