Cloud migration timelines
Channel partners can play a role in cloud deployments and management services to reduce downtime and ensure the migrations go as smoothly as possible.
Across organizations of all sizes, public clouds and private clouds are expected to grow 22% and 15%, respectively, year-over-year for the next two years, while virtual machines and physical machines are expected to decline by 25% and 31%, respectively.
At large companies (with 1,000 or more employees), public clouds and private clouds are expected to grow 52% and 31% year-over-year for the next two years, while virtual machines and physical machines are expected to decrease by 29% and 30%.
42% of respondents use Amazon Web Services (AWS) for current production workloads, followed by VMware (29%). All other platforms account for 29%, of which physical machines account for only 7%.
The top drivers for migration include high availability (27%), reliability (21%), cost savings (21%), and security and compliance (20%).
The top workload types for migration include testing/development (13.7%), Web applications (13.5%), Websites (13%) and databases (13%).
44% of respondents said they can’t afford any downtime during a migration, or at most, just under one hour; 49% have a downtime window of one hour to one day.
The top considerations for choosing a target cloud platform include reliability (22%), high availability (20%), security and compliance (17%) and cost savings (15%).
The biggest migration challenges include minimizing downtime (25.5%), performance impact on production (20.2%), technical challenge/feasibility (19.3%), minimizing loss of data (18.9%) and keeping costs per plan (16%).
55% of respondents estimated the migration cost at $101to $500 per server; 13% estimate the cost at $501 to $1,000 per server; and 4% estimate the cost at more than $1,000.
The execution phase of the migrations lasted, on average, six months, or 45% of a total project’s time. Planning lasted 13 weeks, or 23% of a total project’s time; proof of concept took seven weeks (12%); and test and optimize lasted 12 weeks (20%).