Piszki Lab | EN

My case study in the clouds…

ScaleIO 1.32 installation in vSphere 6.0 – Part Two.

| 2 Comments

In the first part we prepared our environment for the ScaleIO installation. I immediately warn that the procedure of disc drives preparation to work as VMDK is very time-consuming. If we have in each server several large disks (the more the better) in the beginning we can choose the one disk per server. This will speed up the installation, which is of great importance if errors occur and we will have to undo the changes. At a later date, if all goes well, add the remaining disks. It is worth considering the structure of the whole, ScaleIO roughly corresponds to the one known in the disk array. So we have a disk pools to which we assign a specific drives (you do not need to have everything in one pool). Besides, in a similar way ESXi “sees” ScaleIO, SDC component presents itself as HBA (FC).

scale2

We begin by running the “Deploy ScaleIO Environment” option and go through all the configuration screens. It is the first installation, so choose the option “Create new ScaleIO system”. We can also select (at the discretion) to “prepare environment without configure”, this will significantly shorten the time needed to initiate the environment but all configuration settings and so we will have to do later. I suggest for the first time to perform the full procedure.

scaleio10

We give the name of our installation and password for the administrator (admin):

scaleio11

We choose which cluster and which hosts will be part of our ScaleIO installation:

scaleio12

We can choose which hosts will play the roles of Primary MDM, Secondary MDM and Tie-Breaker:

scaleio13

Configure the service “Call Home” and access to syslog (in this case, it is our VMware Log Insight):

scaleio14

We configure our first “Protection Domain”. As you can see, at this stage we set the size of the cache in RAM which will use each of SDS. This value can be changed at a later date.

scaleio28

Within each domain, we can create some storage pools, on this screen define the name of the first pool (or more):

scaleio16

On this screen you can select which hosts install ScaleIO Data Server (choose course all the ESXi).

scaleio17

On this screen for each host individually select which disks will manage ScaleIO:

scaleio18

If you have identical hosts in terms of hardware configuration, you can copy the configuration screen (ie, marked discs) on the other hosts.

scaleio19

This screen might suggest reinstalling the SDC but it is not, on selected hosts SDC is configured (for communication with MDM). These settings can be checked on each ESXi before and after installation ScaleIO (command: cat /etc/vmware/esx.conf | grep scini).

scaleio20

On this screen, select on which ESXi will be installed ScaleIO Gateway, and define a password using the IM logs on to the MDM.

scaleio21

Select the previously loaded SVM template :

scaleio22

We provide management and transfer (for data replication) networks:

scaleio23

On the last screen we define the IP addresses. Here a very important piece of information, documentation misleading and saying to the network management must be added VMkernel. In fact, to the two networks must be added VMkernel (ie Management and Data network). Without VMkernel on data network configuration of the SDC has been executed incorrectly (!!!).

scaleio24

Run the installation process and wait. The installation includes initializing disks (eager zeroed), so it may take a very long time.

scaleio25

In the last step, before moving on to ScaleIO configuration, install the ScaleIO-GUI itself. Here surprise, the GUI installer is included in the “ScaleIO for Windows” and it be downloaded separately (in relation to the version for VMware). Once installed, we launch the console, give the IP address of the Primary MDM (login and password) and we can check the status of our environment:

scaleio27

scaleio26

In the next part we will configure ScaleIO, which will prepare the first volumes and export them to ESXi (and more).

Rate this article:
[Total: 3    Average: 4/5]

Author: Piotr Pisz

Computer always, since I got a Commodore 64 at the end of primary school, through his beloved Amiga and Linux infinite number of consoles, until today, fully virtual day. Since 2001, Unix/Linux Systems Administrator, for seven years a faithful companion and protector of Solaris system, until his sad end. In the year 2011 came in the depths of virtualization, then smoothly ascended into the clouds and continues there today. Professionally working as Systems Architect in the Polish Security Printing Works.

2 Comments

  1. Actually, VMkernel is not needed for the Management network, unless you’re using a single-network option (management + data on the same portgroup). Management network is only used for accessing SVMs. Control and Replication traffic go through Data network. Documentation and deployment wizard are both misleading.

Leave a Reply

Required fields are marked *.


.

Enjoyed the post? Support Piszki Lab | EN, click on the AD! :-)

.