Introducing Atlantis HyperScale

What is it?

A hyper-converged appliance running pre-installed USX software on either XenServer or VMware vSphere and on the hardware of your choice – Lenovo, HP, SuperMicro and Cisco.

How is it installed?

HyperScale comes pre-installed by Atlantis Channel Partners. HyperScale runs exactly the same software as USX, however HyperScale is installed automatically from USB key by the Channel Partner. When it is delivered to your datacenter, it is a simple 5 step process to get the HyperScale appliance ready to use.

Watch the video.

Step 1

Step 2

Step 3

Step 4

Step 5

Done.

What do you get?

The appliance is ready to use in about 30 minutes with three data stores ready for use. You can of course create more volumes and also attach and optimize external storage such as NAS/SAN in addition to the local flash devices that come with the appliance.

Atlantis HyperScale Server Specifications

Server Specifications Per Node	CX-12	CX-24	CX-48 (Phase 2)
Server Compute	Dual Intel E5-2680 v3
Hypervisor	VMware vSphere 5.5 or Citrix XenServer 6.5
Memory	256-512 GB	384-512 GB	TBD
Networking	2x 10GbE & 2x 1GbE
Storage
Local Flash Storage	3x 400GB Intel 3710 SSD	3x 800GB Intel 3710 SSD	TBD
Total All-Flash Effective Capacity (4 Nodes)*	12 TB	24 TB	48 TB
Summary
Failure Tolerance	1 node failure (FTT=1)
Number of Deployed Volumes	3
IOPs per Volume	More than 50,000 IOPs
Latency per Volume	Less than 1ms
Throughput per Volume	More than 210 MB/s

Key Differentiators vs other Hyper-converged Offerings

Apart from lower cost (another post to follow) or you can read this post from Chris Mellor from The Register, HyperScale runs on exactly the same codebase as USX. USX has advanced data services that provide very efficient data reduction and IO acceleration patented technology. For a brief overview of the Data Services please see this video.

Pricing

Sizing

Number of nodes = 4

SSDs per node = 3

SSD capacity = 400GB

Usable capacity per node = 1200GB

Usable capacity per appliance with FTT=1 = 3,600GB

Effective capacity with 70% de-duplication = 12,000GB

Summary

Creating (a better) vSphere 5 ESXi embedded USB Stick (HP)

In a previous post I blogged about creating a vanilla vSphere 5 ESXi USB drive using the VMware .iso file from VMware. This post shows how to create one using the HP version of vSphere ESXi (5.0_Oct_2011_ESXi_HD-USB-SDImgeInstlr_Z7550-00253.iso).

Note: (You can use any vendor customized vSphere ESXi .iso file: VMware, Dell and IBM).

The HP version comes pre-installed with all the HP CIM providers which work very well with HP servers, including the HP MicroServer. Using the HP version gives you the more details in the Hardware Status tab.

I’m going to be using a different method, recommended by Will Rodbard (thanks Will), who is a colleague of mine at VMware, you can see his comments from the previous post. In summary the steps are:

Find and download the following tools:

HPUSBFW & UNETBOOTIN
Run the HPUSBFW tool, click on the USB drive, select ‘Fat32′ and click Format
Run UNETBOOTIN, select Diskimage and browse to the ESXi 5 ISO file
Select the USB drive you have just formatted and click OK
If you want to make more USB keys for more servers, then now is the time to create .IMG files using WinImage, then you can basically clone the image of the USB key to more USB keys. Or if you don’t wish to use WinImage then just perform steps 1 to 4 again.

Once completed your USB drive will boot into the ESXi 5 installer. Once booted, install the ESXi 5 Hypervisor to the USB drive (overwriting the installer). This will then leave you with the installed ESXi Hypervisor on the USB.

Note that using this method creates a brand new bootable USB key for use in a new installation of vSphere ESXi. You will have to go through the process of installing ESXi onto the USB key, or another disk or LUN on the target server. If you want a USB key that is already installed with ESXi which saves you from going through the installation wizard, you can use the other method in this post.

[Aside]

I coincidently left an older USB key in my laptop and booted. Here’s a picture of my Macbook Pro running vSphere ESXi, and it all works by the way, including networking!

Configure NFS Storage on the VMware vCenter Server Appliance

This post highlights some best practices on the management of the vCSA log and core files. VMware recommends that these files are stored on an NFS share external to the vCSA due to the possibility of the default log and core locations filling up.

When this happens, vCenter services will be impacted.

For more information about the vCSA, please see the resources listed here https://vmwire.com/vmware-vcenter-server-virtual-appliance-vcsa/.

There may be trouble ahead

This screenshot shows what happens when this is not done, the partitions for /storage/core will fill up over time and will impact the availability of vCenter Server.

Figure 1 – Local core storage full!

Configuring NFS storage on the vCSA

You can add the NFS shares for the log and core files by logging into the VMware Studio management interface of the vCSA, normally https://<vcsa>:5480.

The default username and password is root | vmware.

Click on the vCenter Server tab, and then click on Storage.

Figure 2 – Configuring NFS storage on the vCSA

Using the correct syntax for the NFS storage

The correct syntax for adding the storage is

<NFS_Server>:<NFS_Export>

So if my NFS_Server is 192.168.200.21 and my NFS_Export is /mnt/vg01/vcsa_core/vcsa_core/, I would enter the following in the box for NFS share for core files:

192.168.200.21:/mnt/vg01/vcsa_core/vcsa_core/

Make sure that the NFS export on the NFS Server is configured with a UID/GID mapping of no_root_squash. For example, use the command on the NFS server:

exportfs -vo rw,no_root_squash,sync :/mnt/vg01/vcsa_core/vcsa_core/

Once done, click on Test Settings to verify that the vCSA can successfully store files to the specified NFS shares, then click on Save Settings, then restart the vCSA.

Browsing to the NFS storage

You can also see what is created in the NFS share if you listed the contents of the core files share.

Figure 3 – Core logs

You can also see what is created in the NFS share if you listed the contents of the log files share. The screenshots below show the directory structure on the NFS server. On the vCSA the directories are mounted at /storage.

Figure 4 – All other Logs

Configuring vCenter Server Virtual Appliance to use an Oracle database

In previous posts I blogged about what the vCenter Server Virtual Appliance (vCSA) is, its features and benefits, feature parity with the Windows vCenter Server and also how to quickly deploy the vCSA. For more information about the vCSA, please see the resources listed here https://vmwire.com/vmware-vcenter-server-virtual-appliance-vcsa/.

This post extends the series with how to configure an external Oracle database for use by the vCSA.

Why use an Oracle database?

The vCSA comes preinstalled with an embedded DB2 database which has similar use cases as the Windows vCenter Server when configured with SQL Express – intended for small deployments of 5 ESX/ESXi servers or less. The ability for the vCSA to utilise an external Oracle database allows customers to scale and manage larger vSphere infrastructures equivalent to environments with Windows vCenter Servers backed by SQL or Oracle databases.

This post shows how quickly and easily it is to use an external Oracle database instead of the embedded DB2 database. Hopefully you’ll see the benefits of how much quicker it is to configure the Oracle connectivity between the vCSA and the Oracle server vs installing the Oracle 64-bit Client onto a Window Server and configuring tnsnames.ora, followed by configuration of ODBC settings.

Configure an Oracle Database and User

Log into SQL*Plus session with the system account. I’m using Oracle 11g R2 x64 on Windows Server 2008.
```
C:`>sqlplus sys/<password> as SYSDBA
```
Run the following SQL commands to create a vCenter Server database. Note that your directory structure may be different.

CREATE SMALLFILE TABLESPACE “VPX” DATAFILE ‘e:/app/oracle/oradata/orcl/vpx01.dbf’ SIZE 1G AUTOEXTEND ON NEXT 10M MAXSIZE UNLIMITED LOGGING EXTENT MANAGEMENT LOCAL SEGMENT SPACE MANAGEMENT AUTO;

Run the following SQL command to create a vCenter Server database user with the correct permissions. I will create a new user named “VPXADMIN” with a password of “oracle”.

CREATE USER "VPXADMIN" PROFILE "DEFAULT" IDENTIFIED BY "oracle" DEFAULT TABLESPACE "VPX" ACCOUNT UNLOCK;
grant connect to VPXADMIN;
grant resource to VPXADMIN;
grant create view to VPXADMIN;
grant create sequence to VPXADMIN; 
grant create table to VPXADMIN; 
grant create materialized view to VPXADMIN;
grant execute on dbms_lock to VPXADMIN;
grant execute on dbms_job to VPXADMIN;
grant select on dba_tablespaces to VPXADMIN;
grant select on dba_temp_files to VPXADMIN;
grant select on dba_data_files to VPXADMIN;
grant unlimited tablespace to VPXADMIN;

Configure the vCSA

Log into the vCSA VMware Studio management interface at https://<vcsa>:5480/
Navigate to the vCenter Server tab, then click on Database.
Select oracle as the Database Type using the drop-down menu and enter your environment information into the fields and then click on Save Settings. Note how easy that was, no messing about with installing the Oracle Client, no need to configure tnsnames.ora and no need for any ODBC configuration either.
Wait for around 5 minutes for the vCSA to create the database schema.
Now it’s safe to start the vCenter services, navigate to the Status tab and click on Start vCenter.
You can then start using vCenter when the Service Status reports as Running.

Cleaning up the Oracle configuration

After you’ve tested that everything is working, you can revoke the following privileges using SQL*Plus again.

revoke select on dba_tablespaces from VPXADMIN;

revoke select on dba_temp_files from VPXADMIN;

revoke select on dba_data_files from VPXADMIN;

Total configuration time ~approx 10 minutes.

References

vSphere Installation and Setup Guide

Creating vSphere 5 ESXi embedded USB Stick

A very quick post on how to create an image that contains vSphere 5 ESXi Embedded with which you can use to quickly create USB sticks that have the ESXi hypervisor installed. This is not the same as creating a bootable USB key that contains the installation files to install ESXi from the USB stick. For this method please refer to this post.

Use this in your lab environment, I wouldn’t recommend doing this in production environments.

In previous versions of vSphere ESXi, it was relatively straight forward to create a bootable USB key which already contained the ESXi hypervizor. This was done by extracting the files from the ISO and then using ‘dd’ to image the directory structure to the USB stick. With vSphere ESXi 5 however, this technique is no longer possible. There is a workaround however. ESXi is installed and configured in two steps, the installation is done to a disk with a vanilla installation of ESXi without configuration. The server is then rebooted and the configuration of ESXi continues with the creation of the management network vmk0 or vmk1 (depending on your setup), hostname, DNS etc.

For this to work, we do not perform the second part, which is the configuration, but take an image of the USB key directly after the installation of the vanilla installation of ESXi without configuration. This enables us to image this vanilla installation onto as many USB sticks, i.e., servers as we like without clashes in virtual MAC addresses and the like.

What you will need: VMware Workstation, 1 USB stick, the ESXi Installable ISO file VMware-VMvisor-Installer-5.0.0-469512.x86_64.iso, WinImage.

Quick steps

Create a new ESX virtual machine in VMware Workstation with CD-ROM drive, USB adapter, 2Gb RAM and 2vCPUs.
Mount the ESXi Installable ISO file to the CD Drive.
Insert the USB stick to your workstation (the same one that runs VMware Workstation).
Boot the VM and connect the USB stick to the VM.
Install ESXi as normal, making sure that you install onto the USB stick, when installation is complete, disconnect the USB stick from the VM and do not reboot the VM, just turn it off. You no longer need this VM.
With the USB stick still connected to your workstation, open up Winimage.
Go to Disk | Creating Virtual Hard Disk image from physical drive and select the USB stick that you installed ESXi on.
Select a location where to save your image and change the file type to Image file (*.ima).
WinImage will now make a backup on your newly installed USB stick.

Creating vSphere 5 ESXi embedded bootable USB sticks

Now that you have an ESXi image, you can use this to build lots of USB sticks which are ready for ESXi deployment.
Insert a new USB stick into a spare USB port.
Launch WinImage and navigate to Disk | Restore Virtual Hard Disk image on physical drive.
Select the USB stick and click on OK.
Navigate to the image file that you created previously. WinImage will now restore the backed up image to your new USB stick.
Repeat as necessary.

Configure ESXi

Once the stick is ready, just insert into a spare USB port on your server and ESXi will boot into the configuration screen ready for you to configure management network details.

You may need to log onto the local console once ESXi has finished booting and launch the ‘Restore Network Settings’. This will reset the vmk0 or vmk1 (depending on your setup) interface.