vCenter Operations Manager - Solving Performance, Capacity and Configuration Problems!!

I have been writing about the cloud infrastructure products of VMware all this while. I believe it is important to look at the management side of things as well. No doubt Virtualization makes things easier for an organization and its IT department; however things are good till the time they are small and easy to manage. As the confidence of such organizations increases towards Virtualization, you would notice a VM sprawl, which might someday defeat the purpose of Virtualization & consolidation.

The result of a sprawl leads to unpredictable behavior of the Virtual Infrastructure, Performance bottlenecks, Waste of expensive resources, complex troubleshooting procedures and issues around other day to day admin activities. I believe any administrator of a medium/large virtual infrastructure would agree to the brief description of the pain points which I have listed above. If you categorize these problems broadly they would fall under 3 Major Categories:-

I - Performance Problems - This will include pain points such as:-

§  How is my overall infrastructure performing?

§  How hard my physical ESXi servers are working?

§  Am I looking at a potential problem which might break my infrastructure?

§  Is there a way I can predict issues and solve them before they are noticed by End Users?

§  Which areas do I look at to troubleshoot any existing issues?

§  Is it a storage issue or is it the hypervisor?

&, the list is never ending..........

II - Capacity Problems - Let's have a look at the pain points of capacity:-

§  Do I have enough Physical capacity (CPU, MEMORY, NETWORK and STORAGE) to support my virtual machines?

§  When do I need to buy more hardware?

§  Am I wasting any resources?

§  Have I sized the virtual machines appropriately?

§  How should I answer questions raised by my CXO's about capacity forecasts, buying decisions etc.??

& many more.....

III - Configuration & Compliance Problems - A more critical problem area, let's see some issues here:-

§  I am asked to follow HIPPA, SOX or PCI compliance policies for my servers, am I compliant?

§  My infrastructure is too big? How do I control changes in my Virtual Infrastructure? 

§  Is there a way to maintain common standards?

The more you dig here... the more issues you would find...

Well, the list above is only a sub set of issues which we face while managing the Virtual Infrastructure. As a reader of this article and an administrator you would be able to add 10 more unique issues to this list when you read this article. However, we still have a bigger issue on hand. I call it the BIGGEST issue. Questions are always raised around:-

- There are so many tools in the market who claim that they can help me with such issues. Which one should I chose??

- What should I do with my existing tools? That's a huge investment which I have already made.

- Can I get a Single Pane of Glass to solve all such issues? (The most common one - People are fascinated with a single pane, I don't know why?)

and another long list of questions.........

I hope you are with me so far and not lost into the issues which you are facing in your Virtual infrastructure because now I am going to tell you how you can solve such issues. I must tell you over here that the solution I am going to talk about has been around for a good number of years, however I have taken my own sweet time to start believing in this solution as it has matured over a period of time. Now that I see this solution working for large enterprises, I guess this is a good time that you can look into this for solving issues related to Performance, Capacity, Configuration and Compliance in your virtual infrastructures.

As the headline of my post suggest, I am talking about vCenter Operations Manager a.k.a. vCOPS. As an introduction, I would say that this VMware solution has been stitched together in the recent past, by plucking out best components from various industry standard tools which have been there in the Industry for a long time. Though there are a number of functions available in this solution, I would talk about the major life-savers here:-

Let me explain each one of them in simpler manner:-

Patented Performance Analysis - A set of 9 patented algorithms which look at performance as a behavior and not a threshold. The engine learns the behavior of your infrastructure by monitoring all the performance metrics. It learns the Normal behavior and only alert you if it observes an abnormal behavior which could lead to a potential problem. Overall gives you the HEALTH of the infrastructure and make you take right decisions in real-time. This was acquired as a part of Integrien acquisition back in 2010.

Purpose Built Capacity Planning & Analysis - This is VMware Capacity IQ which is rolled up into this suite. Those who know the power of capacity IQ would know that it is the only tool available today, which can pin-point at things like, oversized & under-sized VMs, wasted resources, time remaining & capacity remaining to provision new workloads and potential RISKS associated to Capacity of your Physical Infrastructure. It will also help you do tasks such as Capacity Trending & Forecasting for better and accurate buying decisions.

Automated Configuration & Compliance - This ability of the Suite is provided by vCenter Configuration Manager which again is around for a while. It was a part of the IONIX product portfolio, however later it was picked up by VMware from EMC to weave it into vCOPS and complete the entire picture. This is one of the strongest solution which I have witnessed for compliance and configuration management and has the capability of working across virtual and physical infrastructure, across OS platforms and across server architectures.

I hope this gives you some insight on VMware vCenter Operations Manager. I know I am leaving you with a few thoughts around what else this solution can do and how it actually does what it promises to. There would be questions around financial implications and licensing models as well. I will leave you with a few links which will help you learn more about vCOPS and at the same time I will come back with a few more articles which will help you use this solution effectively in your infrastructures.

Before I share those links, here is an interesting fact which might impress you, if you are still not impressed by vCenter Operations Manager:-

The latest version of vCOPS called vCOPS 5.6 (launched at VMworld Barcelona), has the capability to work across multiple hypervisor, multiple cloud platforms (private or public) and allows you to build Self-Healing mechanisms using vCenter Orchestrator as the Workflow Orchestration Engine. Except Capacity Management, this solution can extend into your Non-VMware infrastructure for Performance, Compliance & Configuration Management. (Storage, Networks, Other monitoring tools, HP UX, Solaris, Applications - Exchange, Oracle, SQL, Amazon, Azure, XEN, Hyper-V etc)

I sure see a revolution coming our way.. Get your seat belts on & sit tight :-)

Here are the links which you can use to learn more:- 

VMware vCenter Operations Manager Fundamentals [V5.X] - This free eLearning course covers how to install and configure vCenter Operations Manager as well as how to use its many robust features.

Other Technical Resources and Links - Link to demos, videos, documentation etc.

Pricing & Packaging - All you need to know about vCOPS licensing, pricing etc.

Hope this helps... If you liked this article, kindly share with others and let the knowledge spread.......

Things to Know About Guest IP Customization in VMware Site Recovery Manager!

With the introduction of Site Recovery Manager 5.0, the customization workflows of the product were improved tenfold as compared to the previous versions of SRM. SRM 5.0 Guest IP customization was a great revolution as this helped reducing the RTO significantly. In the earlier versions of SRM, it use to take a long time for Virtual Machines to get a new IP address during a Test Recovery or Actual Recovery since this was achieved by using utilities like sysprep.

However, with SRM 5.0 the Guest IP customization feature allows you to inject a new IP address to the virtual machine powering on at the DR site within a few seconds. This is done by using the guest OS APIs which are used to push the new IP address as soon as the VM is ready to be powered on. In-fact to achieve this, the virtual machine is first briefly powered on to inject this IP and then it will power on again with the new settings as per the power on priorities and dependencies which you have created in the recovery plan.

The amazing thing is that this can be configured for hundreds of virtual machine by using a pre configured xml file or a convenient GUI option can be used, if you have a smaller environment. To learn about how to configure Guest IP customization refer to the This blog article from VMware Blogs.

Now that you understand the feature, its important that you use and apply this in an environment which can support such a feature. The reason I say this is because of the fact that Guest IP customization does not work on all the Guest Oses which are supported on a vSphere Platform. This will only work on Guest OSes which support Guest IP customization. Though, this feature is supported by most of the Guest Operating Systems, however it is better to check before hand to ensure that you do not face any roadblocks during the implementation. You can get the list of Guest OSes which support Guest Customization on the following link.

In case you get into a situation where you have a Guest OS which is not supported for customization, you would get the following error message in the Recovery Workflow as shown in the screenshot below:-

Error: The Guest operating system "oracleLinux64Guest" is not supported. The value in the "quotes" will be the OS which does not support Customization hence if your recovery workflow has the step to customize the IP of this OS, then the recovery plan will STOP with this error. The end state of the Virtual Machine at the DR site would be in registered mode, however it would be in a powered OFF state.

You can manually power on this VM in the DR site, however the IP address of this machine will remain unchanged from the primary site which could result in a catastrophe, hence please be careful.

Now, let's talk about the remedies and how we can take care of such situations till the time the Guest OS advances and starts supporting the customization option. I believe we can 2 methods to take care of this issue.

Method 1 - We can make this change a manual option, which means that we can easily Add a Message Step to the recovery plan for such Virtual Machines. This message step would be added before the Power On step in the workflow. This message will display that the Virtual machine vNic needs to be disconnected at power on and the virtual machine IP address needs to be changed by the Administrator manually at the time of TEST or Actual RECOVERY. You should also disable the IP Customization step for this machine in the workflow so that the workflow executes successfully.

Method 2 - The second method is more ADMIN friendly as I am going to ask you to script this change and add this to the Post Power on script option in the Recovery Workflow. For this you would need 2 Ethernet configuration files, one for Primary and the other for DR site. Please research and create these files on the basis of the operating system which you have. This would mostly be a Unix flavored OS as most of the Windows Guest OS support Customization. Once you have these files ready, add a script on the Startup of the OS to replace the existing network settings with the new one at the time of , Failover, Failback or Test.

Hope this will help you to tackle such a situation if you come across one. 

Using vSphere Replication for Protecting Databases with VMware Site Recovery Manager

I recently came across a question which involved the utilization of vSphere Replication to replicate databases from Protected Site to Recovery Site with VMware Site Recovery Manager as the DR engine. This query had 2 parts to it:- 

• One, whether vSphere Replication Supports DB replication, &
• Is it a good option to use vSphere Replication for DB Protection.

To begin with, you can definitely replicate Virtual Machines using vSphere Replication as a part of the VMware SRM 5.x solution. vSphere Replication does not care about what application you are running inside the Virtual Machine, hence we replicate any and every virtual machine which you configure for replication using this method. This is because we do not do the replication at the VMFS file system layer, however we do this from the VMkernel layer by using a vSCSI filter. 

Now, since we are planning to replicate a database, we might want to consider a few things which might help the replication engine to have a crash consistent data stream on the DR virtual machine. 

Since Databases can always create consistency issues when you replicate them using storage based or host based replication, most of the database vendors have their own solutions around replicating the database at the application level.This is definitely the safest option, along with a more traditional option of log shipping. However today I will discuss about other methods which might prove to be successful in your environments and help you save money on DB replication licensing & management overheads.

"With SRM 5.0 and now with SRM 5.1 we heavily rely on VMware Tools for consistency at the OS and application level.  VMware Tools has the ability to issue commands to the operating system such as to set up VSS snapshots.  With 5.1 we have the ability to do a little more than we have in the past, and ask the OS to flush application writers as well as make the OS itself quiescent.  This means for things like databases, messaging platforms, and other applications that have VSS writers, we can ensure a higher level of application recoverability.  When using vSphere Replication we can flush all the writers for the apps and the OS ensuring data consistency for the image used for recovery." - Reference - Ken Werneburg's SRM 5.1 and vSphere Replication as a Standalone Feature.

The only show stopper for us in this case would be that the OS instance of the Virtual Machine running the Database should support VSS (Windows Only). If your DB is indeed on Windows, then you would be able to enable Quiescing using VSS (Volume Shadow Services), while configuring vSphere replication on a virtual machine. The screenshot below shows that option highlighted in red, click on that drop-down and you would have the option to select VSS.

If you are not on Windows, then you should look at alternate solutions, such as Storage Array Based Replication or Log Shipping. In all the cases, I would setup a test environment and double-check the solution as this is not just technology dependent, but also environment dependent.

Hope this helps you take the right decision.

Best Practices around using RDMs in vSphere!

One of VMware's partner engineer raised this query on an internal group. He wanted to understand and learn the best practices or the Do's and the Don'ts while using RDM (Raw Device Mappings) Luns in a vSphere environment.

I hope you being a reader understand what an RDM is and what role does it play in a vSphere Environment. In case, you are not aware of RDM, then kindly refer to the following document - vSphere 5.x Storage Guide and read about Rae Device Mapping (RDM).

During this discussion we will consider the following requirements which we need to meet :-

• We need to provision RDM's for more than 20 VM's and size of disks will vary from 1TB to 19TB.

• The RDM is decided for configuring MSCS on VMs like MS Exchange, MSSQL, File Servers etc.

A usual topic of discussion is choosing between RDM and VMDK. Since we have already solved that mystery, there is not much to worry about. We are already following the best practices around application layer by choosing RDM’s instead of VMDK. Now since we are playing with Luns mapped to your virtual machines, there are a few things we should take care of:-

1. Choosing between Physical Compatibility Mode & Virtual Compatibility mode for RDM – A physical RDM is more storage array driven and virtual machine controlled. The VMkernel has limited or no role to play and it literally becomes a postman who delivers IOs from the OS to the LUN (just like an OS running on a physical server saving data on a storage LUN). This will restrict you from using VM level snapshots and other file locking technologies of VMKernel. Since you are talking about file sizes of more than 2TB, please ensure you are on VMFS 5 and use Physical compatibility mode only as VMFS 5 does not support RDM with Virtual compatibility mode for Luns greater than 2TB – 512bytes. On the other hand you would be able to use RDM with Physical compatibility mode for up to 64 TB. (VMFS 5 required)
2. Ensure you have the correct Multipathing and failover settings
3. Please ensure you are zoned appropriately. Test heavily before pushing things into production
4. Follow the MSCS on vSphere guide without fail to avoid any last minutes surprises

Well this should help you do the right things with RDM's. Ensure you go through the document which I mentioned before and you should be good to go.

"Target disk UUID validation failed" Error while configuring vSphere Replication on a Pre-seeded VMDK

vSphere Replication gives you the option of performing a host based replication of Virtual Machine from one data-center to another over a network link. This feature was introduced with vCenter Site Recovery Manager 5.x. This allowed customers to use vSphere Replication as the primary replication engine to replicate data from one Site to another and then use the SRM engine to provide automation to the entire DR process.

Since, you have the option to set replication per virtual machine, you can also, pre-seed the VMDK files of a virtual machine on a LUN in the target Datastore (by restoring a full image from a backup). This allows you to save time and replication bandwidth, since you do not have to replicate all the data over the WAN. This will allow you to just replicate the changes from Primary Site to DR Site by Syncing both the images.

Most of the customers, who would use the pre-seeding method would register the restored VM's on the DR site and power them on to check if the backup was good and can they pre-seed that image. Once this VM is registered and powered on, you will be asked a question whether this VM "was copied" or "was moved". If you proceed with the default option of "was copied", the UUID of the VMDKs would change to a random value.

Now when you try to setup the first time Sync using the vSphere Replication configuration wizard, this configuration would fail with the following error "Target disk UUID validation failed".

This error comes up because when the replication engine compares the VMDK descriptor files of Source and Destination VMDK files, they both have different UUIDs. This causes the replication configuration and the first time sync to fail.

To solve this issue, you can simply use the ESXi shell or putty session to get the UUID from the descriptor VMDK from the Primary Site VM. Keep this UUID noted as you would need to replace the UUID of the target VMDK descriptor with this source UUID. Once done, you would be able to setup the Replication again using the same seed vmdk without an issues.

Here is how a UUID would look like in a VMDK descriptor:-

ddb.uuid = "60 00 C2 94 dd 43 63 90-18 77 3f 23 6d 8e f0 22" 

Please ensure you do this for all the disks (vmdks) attached to the Virtual Machine in question. Please ensure you have a backup available before you play around with this, in-case you do not have hands on experience.