2019: The year to depend on “Dependencies”

Based on insights that we have received from our customers, partners, industry analysts we have worked with, and our internal SME team, here are Uila's predictions for 2019. The common theme that we observed, was the importance of up-to-date visibility of “Dependencies” that exist  in Data Centers and Cloud environments. Without this visibility, not only will organizations struggle to ensure top application performance or secure their environment, but also cannot successfully execute strategic transformation projects such as cloud migrations. Here are the predictions in detail:

• Solving Application challenges across cloud boundaries will get tougher

The IT world is definitely seeing the seismic shift towards Hybrid or even Multi-cloud deployments in 2018, and this gets even bigger in 2019. This leads to bigger challenges for IT and Cloud Operations teams to find the exact bottleneck for application outages and performance issues. Without getting into “another silo war” within the organization of where the issue lies, organizations will need to find ways to identify “who” or “what” is at fault and be able to resolve it quickly with configuration changes or applying more cloud or infrastructure resources. Visibility into the dependencies that exist across on-prem infrastructure and the cloud may the only time and cost-effective way to find that needle in this haystack with unified visibility across this new hairball of connections. 

• Security Monitoring for “external” and more importantly “internal” threats

Traditionally organizations have focused and done a reasonable job on perimeter or external threats, but the attack surface has expanded to the multiple assets within the datacenter, including virtual machines, containers, etc.  This “east-west” attack which is based on the land and expand philosophy within the data center calls for the urgent need to reduce the attack surface and secure critical applications and workloads. The first step here is still around full visibility of the infrastructure (physical and virtual), network and application assets operating in the environment, but more importantly on the dependencies that exist between them. Also these dependencies need to be tracked in real-time in this ever-agile world of VMs and containers being spun up, torn down or migrated. A Change Control Monitoring system is critical for organizations to be on top of these dynamic changes in the environment, so that the security policies are always in sync. 

Data Centers need to employ more granular or microsegmented policies to thwart the attack from spreading inside the Data center. With microsegmentation, IT teams assign security policy at the workload level, and that security can persist even when the workload moves across cloud domains. And this can only happen when the organization has full insight into all of the assets and their interdependencies to build that micro-segmentation plan.

• Roll-backs from the Public Cloud continue to grow due to lack of visibility and poor pre-migration planning

Earlier this year, we conducted a survey with VMware vExperts, asking them about their opinion on seeing organizations rolling back from public cloud to on-prem or private cloud deployments.  Almost half of the vExperts see roll-backs happening sometimes, while 33% don’t see it often, but feel it is real. Costs, compliance and unknowns were the biggest reasons for the rollback. Even IDC in a recent survey confirmed that they are seeing a large % of respondents repatriating data back to on-premise private clouds.

Rollbacks are a temporary saving grace to maintain business continuity, but at the end of the day rollbacks need to be avoided in the first place itself. This needs to be done to not only minimize costly, time-consuming rollouts and rollbacks that impact both the organizations’ performance, but also on a personal level, the careers of IT migration and planning teams. 

 Even though unknowns in all situations may not be avoidable, it is important for the IT teams to do everything they can to minimize that likelihood. Planning with good current operations analytics and understanding the current architecture is critical. This can be achieved by conducting pre-migration assessments where you have full visibility into the application and infrastructure assets and their dependencies. The Public Cloud provides the flexibility of scaling up the resources as needed by the application, but a close eye needs to be kept on the expenses as it is very easy to lose track of what the actual spend is with the entire workload or a partial dependency in the Public Cloud. That is why proper planning before the migration is critical to understand the assets involved in the workload and also its current usage for compute and storage. This way you have a good estimate of running the workload in the public cloud, and gives you the early opportunity to either cancel the migration move for that workload or get the right budget (and approvals of course!) when this workload is running in the cloud.  

• Additionally similar to 2018, we still see Virtual Desktops continue their great run, and Cloud is the new arena. 

VDI has continued its forward march with many large scale deployments across multiple verticals, and this is now transitioning to the next level with Desktop as a Service offered by multiple Cloud Providers (AWS Workspaces and the new Microsoft Windows Virtual Desktop for Azure). These cloud-based DaaS solutions open up a new spectrum of option for organizations (led by financial, healthcare and educational vertical) who need to rely on virtual desktops. This will call for VMware and Citrix to evolve their VDI offerings moving forward. But even with this cloud based virtual desktop approach, there will still exist the multi-cloud challenge of finding the problem of desktop outages and performance challenges across the cloud boundaries