VMware notes

【vSphere HA】
When vSphere HA fault damage must occur ESXi host,In the only other surviving HA Cluster ESXi host,Will be stored in the shared storage resources of the virtual host VM restart。

If the underlying ESXi host failure damage incident did not happen,But within vCenter Server Hosting service stops when the operation itself,So vSphere HA high availability mechanisms are not helpful。 (It can make up by Watchdog)

【Watchdog】(6.0More than just support,Automatically enabled by default)
vCenter Server 6.0 adds "Watchdog" monitoring mechanism,When a fault event occurs while the operation of the service stopped working,When the former occurs twice,Watchdog will attempt to restart the service,If the third time is still unable to restart the service,VM virtual host will be restarted。

【參考連結】

——
【VMware Data Recovery】
VMware Data Recovery (vDR) belong to the Disk-based backup,Is simple and easy to deploy features (because it is also a virtual host),Support Windows VSS (Volume Shadow Copy) backup,Also supports simple deduplication function (Data De-duplication)。

However, this backup tool has the following limitations related functions:

·Backup environment must have the vCenter Server,Can use this backup tool。
·Disk-based belong,Or can not use the File Level Block Level backup。
·Each vDR can only back up 100 virtual host。
·Not work on IPv6 network environment。
·Only supports backup operations Windows operating system (Windows 2000 / XP / Vista / 7、Windows Server 2003/2008) Web Hosting,Does not support other operating systems,Such as Linux、Solaris, etc. Web Hosting。

【參考連結】

——
【VMware Fault Tolerance】
Web Hosting have a short downtime VMware HA mechanism to start if the company can not accept,The VMware FT may be a solution to the problem,But Enable VMware FT There are many prerequisites,For example, the virtual host can not take a snapshot (Snapshot)、Stoarge vMotion feature can not be used、The virtual host can only use one vCPU conditions, etc.。

VMware FT mechanism will be on two different were established Primary and Secondary Host Web Hosting,And the use of information technology to VMkernel Port vLockstep on the ESX / ESXi Host to transfer Primary Secondary virtual host to the virtual host,But the Secondary will not write the actual behavior of I / O。

When ESX / ESXi Host Primary fault damage which the virtual host,The Secondary virtual host will immediately take over the relevant job,Primary and become virtual host,At this time will be on another ESX / ESXi Host,Re-establish a new Secondary virtual host to synchronize data with the Primary Web Hosting

【總結】
·vMotion/DRS:This mechanism for "Project of" downtime for,For example when the ESX / ESXi Host entity server occur memory、Hard drive failure,Or when need be shut down for firmware (Firmware) update and annual maintenance,This work is scheduled good plan can use this technique,Will work on the virtualization platform of Web Hosting,Migrate to other stations ESX / ESXi Host,So that enterprises can maintain physical host in the case of service without interruption。

·HA / FT:This mechanism is suitable for "non-program of" downtime。When ESX / ESXi Host server entity without power system problems when power failure,When the host entity or motherboard damage leads to physical host to malfunctions of these non-human factors of non-fault condition occurs plan,Through this mechanism can automatically migrate to the virtual hosts on other stations ESX / ESXi Host to continue the operation of the boot。

But it is important that,These mechanisms are merely protecting ESX / ESXi Host Level level only,Rather than the virtual host operating system level (OS Level),As well as applications on the operating system level (Application Level)。

VMware HA mechanisms such as the previously mentioned,When ESX / ESXi Host fault damage,Equal to its operations on virtual hosts and it was improper shutdown,Although virtual host can be turned on again on another table Host,But it's possible the virtual host operating system has been shut down because of improper operation caused damage to the system,So even if you have migrated to other units on the Host,Success can not be a smooth start,So the backup job its necessity operating system。

The VMware FT mechanism is to make two virtual hosts operate exactly the same data,Therefore, if the situation when the crash occurred Primary Web Hosting,For example, the Windows operating system crash occurs blue screen (Blue Screen Of Death,BSOD),Because this time will vLockstep synchronization mechanisms,Secondary and makes the virtual machine host system when the situation also occurred。

As for application-level protection mechanisms,Currently, there are many manufacturers develop mechanisms,For example Application HA Symantec with Veritas Cluster technology development,Is to protect the application to run on a virtual host,Such as MSSQL、Exchange、Oracle、SAP, etc.。

【參考連結】

—–
【VMware DRS】
The auto-load the host entity , Automatic scheduling virtual location system , That is, when a higher host load station entity , Through DRS can be part of the subsystem through VMotion to move to a lightly loaded host entities。

Leave a Comment

Please note: Comment moderation is enabled and may delay your comment. There is no need to resubmit your comment.