Tag Archives: vmware

VMware Data Center Virtualization Fundamentals – Part 7

In this post we will continue our look at the VMware software, specifically vSphere 5.5.

  • Virtual Machines concepts
  • ESXi Hypervisor
  • vCenter and its features
  • Storage and Virtual Networking
  • vCenter Operations Manager
  • vCenter Configuration Manager
  • vCenter Site Recovery Manager

So far we have looked at the components of the vSphere system, and the handy chart below shows which editions come with which features but there are some extra pieces that we should also look at.

vsphere-5.5-editions

vCenter Operations Manager

vCenter operations manager monitors your vSphere system for trends that could cause problems. It looks at aspects of performance and capacity and puts these into three categories:

  • Health – immediate issues that need attention now
  • Risk – future performance and capacity issues
  • Efficiency – right sizing VM CPU, Disk and memory issues

The thresholds for these are dynamically configurable and can be adjusted to prevent false positives. Current capacity can also be monitored (real time) and there are a number of editions available, depending on your business model and requirements.

vCenter Configuration Manager

The vCenter configuration manager can:

  • Track changes in configuration and correlate them to issues or problems.
  • Offers configuration analysis; allowing a compare to third party specs for security hardening for instance.
  • Automate configuration compliance
  • Helps keep regulatory compliance

All of these features can trip alerts and automated remedies which is in part what helps maintain regulatory compliance and security. Any configuration changes are tracked and monitored, allowing post mortem corrective actions if required.

vCenter Site Recovery Manager

SRM is a disaster recovery tool. It makes fail over and testing of disaster recovery plans simple. In addition, it can replace a paper plan with an automated plan. In addition, SRM can perform frequent non disruptive testing of that plan.

SRM can also be used for data center migrations. SRM uses vSphere replication to replicate specific VMs or all VMs – it is all configurable.

 

 

 

Tweet about this on TwitterShare on Google+Share on LinkedInShare on FacebookShare on RedditShare on StumbleUponEmail this to someone

VMware Data Center Virtualization Fundamentals – Part 6

In this post we will continue our look at the VMware software, specifically vSphere 5.5.

  • Virtual Machines concepts
  • ESXi Hypervisor
  • vCenter and its features
  • Storage and Virtual Networking
  • vCenter Operations Manager
  • vCenter Configuration Manager
  • vCenter Site Recovery Manager

Storage and Virtual Networking

We have seen from the previous posts, how the VMs get their memory and CPU from the ESXi host, but what about storage?

While it is possible to run ESXi on a single host with a local storage drive, typically with multiple ESXi hosts, the storage is shared storage. This allows some of the advanced features we discussed in previous posts to operate. Moving VMs between hosts, requires shared storage.

6194360-591-517

These services all require shared storage:

  • DRS
  • DPM
  • Storage DRS
  • HA
  • FT

So what types of shared storage can we use:

  • Fiber Channel (FC)
  • Fiber Channel over Ethernet (FCoE)
  • iSCSI
  • NFS

… and of course, as I stated earlier, we can also use local storage, but that limits our capabilities. However there is a part of vSphere called VSA that allows for some of that capability even on local storage.

Virtual Storage Applicance

VSA allows us to create virtual shared storage on local storage.

VSA

VSA allows for replication between virtual drives, and is perfect for SOHO applications, keeping cost low.

While we are on the subject of local storage, we can also use vFlash. vFlash acts as a cache for VMs, making read speeds much faster than from mechanical drives. This requires at least one SSD drive in order to work.

Virtual machines are stored on these drives in containers called Datastores. These are essentially walled gardens that hold the VM. We can open the datastores and see the files inside them, even upload and download to them.

The VMs can’t see the underlying storage scheme; to the VMs, it looks like they are stored on SCSI devices, which is what makes it easy for VMs to be moved, replicated and copied. However, there are two types of datastore; these are VMFS and NFS. The VMFS type are used for local, iSCSI, FC types of storage where are NFS is used for NAS stores.

Virtual networking

Our virtual data center has virtual machines, and virtual storage. Both of these require networking and some of that networking is virtual as well.

We use Ethernet for

  • IP Storage – iSCSI, NFS, FCoE
  • vCenter Management – vMotion, HA, DRS etc
  • VM Communication – network traffic to other VMs and the real world

vnet

In the ESXi Server, each VM has its own virtual NIC connected to a virtual switch, that connects to the real NICs. All of this is of course configurable.

There are two types of virtual switch available:

  • Standard Virtual Switch Architecture – manages VM and networking at the host level
  • Distributed Virtual Switch Architecture – manages VM and networking at the data center level

Features of virtual networking include:

  • VLANs
  • Traffic Shaping
  • Port Mirroring
  • QoS & DSCP
  • CDP & LLDP

 

 

Tweet about this on TwitterShare on Google+Share on LinkedInShare on FacebookShare on RedditShare on StumbleUponEmail this to someone

VMware Data Center Virtualization Fundamentals – Part 5

In this post we will continue our look at the VMware software, specifically vSphere 5.5.

  • Virtual Machines concepts
  • ESXi Hypervisor
  • vCenter and its features
  • Storage and Virtual Networking
  • vCenter Operations Manager
  • vCenter Configuration Manager
  • vCenter Site Recovery Manager

vMotion

The vMotion is the center piece of the vCenter software. It allows the migration of VMs from one ESXi host to another without interruption of service. This means that server maintenance can be performed during work hours negating the need for overnight or weekend shutdowns.

But it is more than that because it also allows for the migration of VM(s) before hardware failure, thus essentially eliminating downtime due to hardware failures.

But wait, there is more! It also has the ability to load balance, and it can do all of this without shared storage. The key is that virtual machines (mostly) exist in memory and vMotion handles the replication of memory and transfer of control.

Distributed Resource Scheduler

The DRS is a vMotion automation system. This system does a number of things:

  • Monitors the memory and CPU load on ESXi Servers
    • unevenly over burdened ESXi hosts can result in hotspots
    • hotspots will impact VM performance
  • DRS can balance VM load by using vMotion to move VMs across multiple hosts
  • Follows ‘rules’ allowing segregation of VMs
  • Shut down VMs that are idle during off peak times

Distributed Power Management

DPS is used to provide shut down and start up services for VMs. In conjunction with DRS, it can

  • shut down idle VM
  • Start up VMs as load increases
  • Consolidate VMs during idle times

As server load increases, DPM can power on more servers to provide more resource for VMs. DRS and DPM together balance power consumption and availability of services.

Storage vMotion

Does the same thing that vMotion does, but does it for Hard Disks and storage devices. It monitors and controls storage and allows the migration of data from one storage device to another seamless to the user. This allows for hot upgrades.

Storage DRS

As with DRS, there is a Storage DRS also. Using storage DRS allows the monitoring and load balancing of storage devices. We can cluster storage devices and let Storage DRS handle the load balancing (with Storage vMotion). Storage DRS will also tell us if a particular storage device is too full or over utilized.

Storage IO Control

Really busy servers, like email and data base servers tend to consume a great deal of IO resource. This can negatively impact other VMs that use the same storage. Storage IO control essentially caps IO access for high IO VMs, leaving some room for the other VMs in the cluster.

vSphere Data Protection (VDP)

This is the mechanism by which we back up our virtual machines. VMs use image based backups, otherwise known as bare metal backups. This is a disk level image of the VMs and can be stored as a file. Because this is file based, it is easy to make backups of VMs, and simple to restore them. VDP uses data duplication, and is a file based backup so we can restore individual files as well as the complete virtual machine.

VDP is bundled with most version of vCenter and can be run from the vSphere web client, however the APIs are open (VADP) and allow for the use of third party backup tools.

High Availability (HA)

So what happens when the actual host hardware dies on you? Each ESXi instance communicates with other ESXi instances in its cluster. When a host goes down, the ESXi instances detect this and attempt to start the failed VMs. The usual down time for a VM in this case is less than one minute.

But what if the VM itself goes down but the host is still up?
Each VM has a special set of drivers and software that monitor its own operation – like a watchdog. This watchdog tells the ESXi software that the VM is working ok. if this changes, ESXi will restart the VM. This watchdog will also monitor applications, so if an application becomes unstable and fails, HA will let you know and if configured, will attempt a restart.

Fault Tolerance (FT)

Some servers or services cannot be offline and cannot go down under any circumstances, because of data loss or corruption. In this case, vCenter has a service called Fault Tolerance or FT. This allows for the mirroring of VMs across multiple ESXi hosts. If one VM or host goes down, another will automatically take up the load seamlessly (co-execution)

vSphere Replication

This service allows the replication of VMs to another ESXi cluster at a remote location. This is useful for disaster recovery. Replicate your data center to another building, city or state, so that in the event of a natural or other disaster. your data center is protected. You can also do the reverse in that branch offices can replicate their VMsto the central data center to act as a backup and as a means of data protection.

Tweet about this on TwitterShare on Google+Share on LinkedInShare on FacebookShare on RedditShare on StumbleUponEmail this to someone

VMware Data Center Virtualization Fundamentals – Part 4

Continuing from the last post, we will continue our look at the VMware software, specifically vSphere 5.5.

  1. Virtual Machines concepts
  2. ESXi Hypervisor
  3. vCenter and its features
  4. Storage and Virtual Networking
  5. vCenter Operations Manager
  6. vCenter Configuration Manager
  7. vCenter Site Recovery Manager

3. vCenter and its features.

As we mentioned in the last post, we need a central management system that is easy to use, if we are to manage multiple servers, with multiple VMs on them. This is where vCenter comes in.

vmw-dgrm-vcenter-server1-lgVMware’s vCenter is a group of tools that enable to user to run a single ESXi instance to a whole data center full of ESXi instances. The software is scaleable and flexible, including:

  • Identity management Server
  • Data base Server
  • Application Server
  • Web Server
  • VMware vSphere Web Client

From this console, we can run up to 1000 ESXi hosts, and up to 10,000 VMs. From here, we can add more instances of vCenter, to expand that capacity, and include cloud based management.

In addition there are some other tools like:

  • vMotion – that allows the migration of a running VM from one ESXi host to another without interruption of service.
  • Distributed Resource Scheduler (DRS) – provides for load balancing across ESXi hosts. With vMotion, this allows for multiple VM instances sharing workload spread across multiple servers.
  • Distributed Power Management (DPM) – is used to power off unused ESXi hosts, and power them back up when needed based on load or other criteria.
  • Storage vMotion – allows the migration of VMs hard drives across multiple ESXi hosts
  • Storage DRS – does the same load balancing for the hard drives
  • vSphere Data Protection – backs up virtual machines. This also includes:
  • HA – High Availability, which gives allows the migration of VMs in case of hardware failure, along with
  • FT – Fault Tolerance which provides a more seamless transition between machines if the start up period to get the VM running is too much outage.
  • vSphere Replication – can copy VMs to another site for disaster recovery purposes

In my next posts, we will look in more detail at these tools.

 

Tweet about this on TwitterShare on Google+Share on LinkedInShare on FacebookShare on RedditShare on StumbleUponEmail this to someone

VMware Data Center Virtualization Fundamentals – Part 3

In this post we will look at the VMware software, specifically vSphere 5.5. This software can be broken down as follows:

  1. Virtual Machines concepts
  2. ESXi Hypervisor
  3. vCenter and its features
  4. Storage and Virtual Networking
  5. vCenter Operations Manager
  6. vCenter Configuration Manager
  7. vCenter Site Recovery Manager

1. Virtual Machine Concepts

So just what is a virtual machine?

  • Looks like a physical server to the user
  • Users can’t tell the difference
  • Virtual but dedicated hardware for each OS instance

esx_io_2

The advantages of virtual hardware are:

  • Replace and upgrade (hardware) on the fly – no service interruption
  • Add new devices without rebooting
  • Stable and dependable
  • No need for OS modifications

Using the vSphere web client we can manage servers remotely, including OS installs from CD/DVDs or ISO images.

What can we do with a virtual Machine?

Given that a virtual machine behaves just like a physical machine, we can almost anything a physical machine can do including:

  • File Server
  • Data base server
  • Email server
  • Applications server
  • and many other server types or applications

So why are we so interested in virtual servers? There are a number of reasons why we might consider running virtual servers:

  • Consolidate a number of lightly used servers – to conserve space and power
  • Protect against hardware failures/site disaster – VMs are easy and quick to copy, backup and replicate
  • Ease of provisioning servers – quick to set up and deploy new servers

There are some things that virtual machines can do that a physical machine cannot do.

Snapshots are images of the running virtual server. They can be used to ‘roll back’ any issues or problems with the server; perhaps caused by a bad driver update or patch. This can be achieved in minutes, as opposed to a complete server rebuild, which can take hours.

It is important to note that snapshots are NOT backups of the server. For a complete backup, we use the Image backup. This is a similar process that allows for backup of the entire ‘metal level’ of the server, and allows for file based restore.

Both of these types of backup have another advantage in that the VM is not tied to specific hardware. In the event of disaster, you may not have access to specific pieces of hardware, so having VMs means that you can recover from disaster on what ever hardware you have available if necessary.

In addition, we can add server resources (like memory, disk space and CPUs) with a reboot. With a traditional hardware server, it must be shut down and the case opened to add resources, but with a virtual machine, we can simply allocate more resources simply, and without having to reboot. This means that maintenance can be done on the fly.

2. The ESXi Hypervisor

As we mentioned previously, the hypervisor is the resource allocation system that interfaces the virtual machines to the underlying hardware. The hypervisor is also our management system to control the VMs. We issue instructions to the Hypervisor that control the machines and the resources allocated to them.

Before we go any further, it should be noted that there are two types of hypervisor:

  • Type 1 -Bare Metal
  • Type 2 – Hosted

virtual-linux-figure1

In this course we are focusing on the type 1 – bare metal. This means the Hypervisor is installed directly to the server without an operating system and OS’s are installed on the virtual machines. In the case of WMware, this is the ESXi Hypervisor.

The type 2 – hosted hypervisor is the type I have on my iMac (I’m running Xubuntu Linux 12.04 with Oracle’s Virtual box) and I use it to run a number of headless Linux server instances and for those legacy programs (read old games) I run windows 98, XP and Win7 as well as an instance of Server 2003 R2 – but not all at the same time! The VMware hosted hypervisor is VMware Workstation.

Bare metal types are the preferred type for data center installations, as they are not dependent on a host OS and generally consume less machine resource. Hosted types are generally used in simple installs for legacy software support etc on a users workstation, rather than a server (except maybe in SOHO applications where cost is more of an issue than performance.

ESXi can be installed from CD/DVD direct to the target server hard drive(s) but also to SD cards, flash cards and USB drives. It can also network boot PXE and TFTP servers.

Using the ESXi Hypervisor, we can:

  • create virtual machines
  • adjust virtual machine configurations
  • monitor performance
  • configure and patch hosts

… which is great if we have only one piece of hardware running ESXI with some VMs. However, if we have multiple servers, hosting multiple VMs, we need a better way to centrally manage them. This is where vCenter comes in.

 

Tweet about this on TwitterShare on Google+Share on LinkedInShare on FacebookShare on RedditShare on StumbleUponEmail this to someone