Overall Satisfaction with IBM Turbonomic
Virtual environments can be challenging to manage. There are only so many resources on a host, and ensuring that each virtual machine has enough to properly function without choking the physical host can be a difficult balance. While some vendors such as VMware provide their own solution, these can be costly and will only work with their products. What does a company do when they have a heterogeneous environment that may also include Hyper-V and XenServer? A heterogeneous virtual environment is one if the reason Turbonomic was installed in my environment earlier this year. This is a powerful product that we have just scratch the surface on its vast features. It monitors the entire virtual environment from the application to the storage pool and provides statistics and recommendations. The objection of Turbonomic is to reach an ideal balance of host resources and virtual machine performance. Once recommendations are made, the system can be configured to answer these adjustments in a variety of ways such as fully or partially automated, manual, or simply recommend.
- In fully automated mode, the virtual machine may automatically have additional memory or processors added, be migrated between host or storage, or even suspended if they system is deemed as idle. If there is hesitation around allowing a product to fully control the virtual environment, then a mix of partial automation and recommendation may be the desired configuration. As an example, the configuration may allow for an automated relocation of virtual machines onto different host, but only recommend that it be migrated from one datastore to another one. This give an administrator who is intimately familiar with the environment some control over the placement of resources.
- Using the Supply Chain tool, an individual virtual machine can be analyzed to see more details on the issue with that particular system. When that machine is selected, utilization panel can be adjust to only show the impact on memory, CPU, storage, etc. There are also details given on where in each resource falls within their utilization index, and at what level Turbonomic has determined the optimal usage should be based on the recommended changes to the environment.
- Virtual environments have made planning for resources rather difficult over the years. Many times newly added machines may take up more resources than previously estimated, causing the need to either halt projects or unexpectedly purchase additional hardware for the host environment. VMTurbo has a planning feature to better aid in determining how many new virtual machines can be added to the environment. This is done via a “template catalog” which is used for testing the environment. While there are some pre-canned system options such as small, medium, and large, there is also the ability to build a custom template that may be a standard for your environment. Once the template is built, the planning tool will tell you how many more virtual machines can be added to the existing environment, the projection option can be enabled and the furthest date out selected. When the report is complete, the future resource requirements will be displayed. This allows for better budgetary planning within the coming year to ensure the systems will perform to the expectations of the business.
- They have expanded the capabilities of the product to now include support for virtual machines running in cloud service providers. This provides a "single pane of glass" view into all of the virtual machines, and their performance, for an organization. Without it, there would be a need to log into various portals to look and analyze performance across the environment.
- They have moved their model from an per CPU pricing model to a per virtual machine subscription model if you want to be able to move to the latest and greatest versions. With many organizations moving to cloud, this is understandable, one of the primary reasons for the original purchase was to move off of vCOPs per virtual machine licensing model to a straight per CPU model. This prevented the need for budgeting due to increasing numbers of virtual machines throughout the year.
- The newest models do not always include monitoring of all types of storage arrays. We moved to a DellEMC Unity about 5 years ago, and as of our last renewal, they still did not have that as a supported array for the storage module. Therefore, we dropped the licensing for this module.
- Turbonomic has provided the infrastructure team better insight into what can be done to improve the performance of our virtual machines. In the past, this could be more difficult to determine. It has also provided solutions where limitation had been set in the past and forgotten about. These settings caused other limitations despite adding more resources. Had we not had Turbonomic these limitation would have either taken more time to location or could have resulted in a support call with VMware.
- We now have a central location to manage each of the virtual environments. While something like vCenter Operations Manager works great with vSphere, if there are other products in the environment, you would then need multiple monitoring solutions. This increase licensing costs as well as complexity.
- VMware Operations Manager
Unlike VMware Operations Manager, Turbonomic has the ability to monitor and manage other virtual environments. This provides the support team a central management point for each environment. It also provides planning capability for each one of these based on the average virtual machine. By creating a template for each environment, future planning can be group based and more accurate than the manual estimate previously used.
Do you think IBM Turbonomic delivers good value for the price?
Yes
Are you happy with IBM Turbonomic's feature set?
Yes
Did IBM Turbonomic live up to sales and marketing promises?
Yes
Did implementation of IBM Turbonomic go as expected?
Yes
Would you buy IBM Turbonomic again?
Yes
Having originally inherited the management of the virtual environment, installing Turbonomic addressed a number of performance issues that were due to settings by the previous administrator. There were virtual machines where reservations or limits has been set on a number of virtual machines. This has never been something I had experienced in an environment and would not have thought about checking this on some virtual machines. The settings were rather random and there seemed to be no sense to the settings. Maybe they had been set as part of troubleshooting and issue and not been reversed. Either way the recommendations to remove those settings was very helpful. The product also monitors and makes recommendations on increasing or decreasing resources on virtual machines such as CPU and memory. This can be set up to be performed dynamically or as recommendations. Either way, it helps ensure proper performance of the environment.
Performance problems can be difficult to narrow down. Especially, if an application owner comes and says they saw an issue in their application in the middle of the night or over a weekend when looking back at logs can be very time consuming. With Turbonomic, there is an easy way to look back at how the virtual machine(s) were performing at a particular time. It keeps a running log of performance in a database. This way if there were issues around CPU, memory, or storage that shows up during the reported window, adjustments can be made quick to avoid a future problem. In the end, it save a lot of troubleshooting and analysis time.
Having the ability to report on capacity and project capacity planning has been helpful in reducing the need to "guess" at needed resources. For example, when looking at performing a data center refresh for the UCS blade environment, the configuration of the newest blades can be used to create a profile. This can be applied to the currently running environment and generate a report on how many of these blades would be needed. This prevents the guess work of just replacing all of the existing blades, or adding additional one for "good measure". While there has not been any move for use to infrastructure in a cloud provider yet, it is coming and this same process should prove beneficial for that environment as well.
Using IBM Turbonomic
4 - This number composes the entire server team. We are using this product for the management of our heterogeneous virtual environment. It not only helps us ensure an optimal performing environment, but will also aid us in planning on the necessary resources going forward. Moving forward, the procurement of hardware assets will be more accurate.
4 - The primary skill would be an experience in the management of vSphere. If there are other hypervisors in the environment, a working knowledge of those systems should also be known. Since this product also reaches into storage, a working knowledge of the storage product and configuration is also helpful. This enables a better understanding of the information coming out of Turbonomic. There is also the need to do some customer configuration on the environment such as storage IOPS in order to ensure the information being reported is accurate.
- Central management of all hypervisor environments
- Future planning of needed resources
- Ensuring optimal performance of the virtual environment
- Ability to monitor the overall utilization of the storage environment
- We are still relatively new to the product and have not moved into any creative uses for it.
- There is an application performance module that may be purchased to ensure the performance level of mission critical applications.
- There is the possibility we may deploy virtual desktops in the future. Turbonomic could then be used to ensure the performance of the virtual desktop environment.
- This product may also be used for monitoring and control of our UCS environment.
Evaluating IBM Turbonomic and Competitors
Yes - Previously, we utilized VMware Operations Manager to monitor our VMware environment. There were two problems with this product. The first was the licensing cost. Besides purchasing the product, there was an additional charge for each virtual machine we had in the environment for it to monitor. The second problem was that it was not the only hypervisor we utilized. Therefore, we wanted a product that was less complicated and expensive, but that also provided a central management point for all of our hypervisors. Therefore, we selected Turbonomic.
Rather than being licensed based on each virtual machine, Turbonomic is licensed based on the number of processors in the environment. The allows the annual cost for maintenance to remain relatively flat. The product offers a wide variety of features that not only allow for better monitoring and performance adjustments, but also have the ability to plan for future use and consumption. However, the depth the product goes can mean it will take some time to get it fully configured to the environment. While defaults are built in, these may not be accurate depending on the size of the environment. For example, the default IOPS provided may be less than what the current system can handle. This is especially true if you have a large SAN or NAS array.
After using this product for about twelve months, I can say we would purchase the product again. The installation process was very simple and the level of support provided by VMTurbo has been great. New updates and releases come every quarter and provide not only "bug" fixes, but new features and capabilities. The upgrade process is performed through the web interface and takes approximately ten minutes to complete.
IBM Turbonomic Implementation
- Implemented in-house
Change management was minimal - The the product was non-user impacting, we did not involve any real change management other than letting people know we were doing it. Just as a FYI.
- No issues
IBM Turbonomic Support
Pros | Cons |
---|---|
Quick Resolution Good followup Knowledgeable team Problems get solved Kept well informed No escalation required Immediate help available Support understands my problem Support cares about my success Quick Initial Response | None |
Not Available
Yes - I ran into a bug with the storage module when I upgraded my version. Support worked to address it and worked to roll back my version until they had a new version available where the problem was corrected.
I was due an upgrade after having ran into issues with a previous upgrade. They basically had about 5 people on a support call for the next upgrade to ensure my next one went smooth. Turbonomic's support has always been great.
Using IBM Turbonomic
Pros | Cons |
---|---|
Like to use Relatively simple Easy to use Technical support not required Well integrated Consistent Quick to learn Convenient Feel confident using Familiar | None |
- The reporting letting me know about changes to virtual machine need either more or less CPU\Memory. This prevents a lot of over or under provisioning of these costly resources. The system can be configured to automatically make these changes, but we are a very conservative company when it comes to risks, so we do not allow the automated function.
- The monthly reports that can be configured to email to the group on a designated date and time. The allows the team to quickly review several of these to see if we need to make changes.
- While not necessarily cumbersome, configuring all of the policies can take some time. The product is so versatile that if you have many different virtual environments, policies have to be created under each one. For example, if you have multiple instances of vCenter, Hyper-V, or XenCenter or any combination, you need to create policies for each one. You could set one over all, but I am not aware of any environment that uses the same policies on every single virtual machine in their environment.