Cloud Enablement “I chat with my infrastructure!”

screen-shot-2016-11-02-at-01-53-38With Tintri you have the ability to automated IT processes with Slack (a chat program, if you don’t know it). From Slack you can create, delete, clone, snapshot a VM, recover a VM, just by chatting with a Tintri Bot that calls our VM-level API engine (and VMware API), and takes care of all the processes in the background.

Sound crazy, maybe. But it’s a fast and easy way of building a simple Private Cloud solution. So the IT department can focus on the bigger things and deliver faster and better services to the business.

Lets say that someone from the application team wants a new VM, instead of they have to contact the VM admin and wait, they can just open their Slack program and write a message to TBOT (Tintri Bot):

TBOT will then create a VM in VMware, as a VM-admin you don’t have to do anything. The application owner will then like, to create a snapshot the of VM, before they update mysql server with a new patch. Then they just write a message to the TBOT:

The application owner is now ready, to set the application in production. To make sure that the VM gets the right storage resources, the VM needs to be a member of the production tier (Tintri then guarantee X number of IOPS per VM, based on the Service Group policy you set). Like before, just chat to the TBOT:

Tintri VM Scale-out and VM-level QoS will make sure that every single VM get the performance it needs in every datastore, no task required. The Infrastructure people just need to look at the hole Tintri pool resources.

Is there free performance? Yes. Is there free capacity? Yes. All good then 🙂

But what if there is a bottleneck somewhere else in the infrastructure? With Tintri VM-aware Storage you can see performance bottleneck in the Network, Hosts and Storage layer for every single VM in realtime. In Secs you know where the issue is and can start fixing it.

That is the power of Tintri VM-aware Storage, and you can start small and scale up to 10 PB and 160.000 VMs in a single solution.

Do you want to know more? Please contact me directly uslothuus@tintri.com / +45 3131 9891

Some IT jobs will disappear…

New ways of building datacenters is starting to kickoff. I read a survey based on 1300 Enterprise customers globally, the outcome was that 60% of them will consider deployment of Software Defined Storage (SDS) and/or Hyper-converged Infrastructure (HCI) in 2016.

If I was a storage admin today, I will start educating myself to other IT roles. Because with SDS and/or HCI there will be no need for storage admin in the data centers. VM and Application admins get full control on where and how data shall be stored, all based on IP, no more SAN zoning / no LUNs etc.

The trend I have seen with my customers is that they don’t have any storage admins, they have been re-educated to VM admins or other roles.

Another interesting point in the survey, was that 20% will also consider Docker, that will properly also change the IT roles. With Docker, applications runs in containers with a set of binaries, very fast to deploy and easy to scale. So if more and more applications will run in Docker containers, without any need of a VM with OS, that will change the VM admins role in the data center as well.

The future data center will be a fully automated and self-tuning Application Store for the business units. The IT department role will be to build that, and then make sure to add new applications if users requires it. Companies will properly need less IT admins and more analytics, developers etc. in the next 5 years.

 

TCO calculation tool for storage

Screen Shot 2015-07-23 at 11.34.22

 

Time to highlight the “hidden” cost!

I have created a TCO calculation tool that can help highlighting where the cost is over 5 years. Reports from my customers and the market is showing that more that 50% comes fra housing, power, cooling and management.

C-level management shall not only look that the purchase price for the storage system, but also look at the “hidden” costs.

New smart application-aware storage systems can cut 50% on the cost on power and cooling over 5 years, good for your budget and the data center will a more greener profile.

About the management part, a smart application-aware storage system (like Tintri) does barely not require any management at all. A smart storage system shall just be one big data store (no management of LUNs, RAID and Volumes), self auto tuning and be install in just 15 min. If there is a performance issue it shall be possible to detect the bottleneck in just a few seconds. That is want I will call smart storage :-) Some of my customers has cut the management cost from 19% of the TCO, to just 3%.

The prices in the TCO tool is just numbers, you have to change all the grey cells with your numbers.

Remember to give the vendors the same requirements with capacity and IOPS over 5 years. IOPS can be a challenge, because there is a lot of factors to play with and it’s easy to make the result better that real-life experience.

You can tell the vendors that you need: 50.000 IOPS (not peak performance), with 8 KB block size, 70/30% reads & writes, 100% random IO. Then make sure that they also can handle your future growth on both capacity and IOPS (put these requirements into your RFP document). There are vendors that are over-selling their performance, and it can be a painfull and expensive relationship over 5 years. I have experienced that myself, when I was a storage admin at TDC.

Link: Storage TCO, 5 years

Data centers in 2015-16

The strategy on how to build effective Data centers will change this year (2015). More or less everything will be virtualized, that is not new. But it will give us the option to start using Smart Application Aware Storage with:

– No need to think about RAID groups, LUNs and volumes, just one big datastore in the hypervisor.

- Selftuning filesystem, that works per VM, instead for LUN and Volumes.

– Granular VM snapshots and replication.

– VM level QoS & analytics/reporting.

All to lower the TCO 🙂

The backup / recovery strategy will change from the traditional streaming backup to snapshot and replication based, archiving VM and data will be needed to break the growth rate.

Mobility and flexibility is a demand from the end-users, their requirement to IT department will be that they can work anywhere, anytime and from any device (BYOD). To be able to achieve this, IT need a single data management platform with snapshot management, archiving, content indexing capability and with a “Dropbox” functionality to the end-users smart devices.