Load profile
For a week we record CPU, memory and disk operations on your current servers, then size cores, RAM and space with three years of growth in mind.
We work from load figures, not a hardware catalogue, so the host ends up neither underpowered nor overpriced.
For a week we record CPU, memory and disk operations on your current servers, then size cores, RAM and space with three years of growth in mind.
Local NVMe, ZFS, Ceph or a shared array: the right answer depends on how many hosts you have and how fast you must be back after a fault.
Separate VLANs for management, guest traffic and replication, so backup streams never throttle your users.
Physical machines become VMs one by one, each in a short after-hours slot.
Ready images of Windows Server and Linux, patched and carrying the monitoring agent. A fresh test server comes from a template, with no install from scratch.
Replication to a second host or the cloud, and a written runbook stating what gets started first when a host goes silent.
Project length depends on the number of servers and on hardware delivery, which you arrange with your own supplier.
A week of data gathering and a conversation about which systems are critical and which can be switched off.
An exact host configuration to order from your own supplier, who installs it on your premises or at the data centre.
Once power and network are connected, we log in via iDRAC or iLO, install the hypervisor and networking, and benchmark it.
Servers migrate singly, users confirm all is well, and only then is the old hardware powered down.
A host with no remote console is a host somebody has to stand beside. When you buy one, insist on an iDRAC Enterprise, iLO Advanced or similar licence and get the management port plugged into the network. Then we can power it up, enter the BIOS and install the OS without anyone's help, and your staff are needed at most to swap a disk.
Yes, that is normal practice now. The one condition is that the SQL Server database gets quick disks and reserved memory rather than fighting the file server and backup VM for them. Before migrating we time routine operations, such as posting a document, and compare afterwards.
Licensing depends on the edition and on the number of cores in the host, and editions differ in how many VMs they cover. A handful of servers usually calls for a different option than a dozen, and Microsoft revises its terms from time to time. We do the sums on current rules before purchase, as an error can be expensive.
With a pair of replicated hosts we boot the copies on the survivor, losing no more than the last replication interval. With one host we rebuild from backup onto spare kit or into the cloud, which takes considerably longer. The right choice depends on what a lost day costs your business.
Often, not always. We model three years of licence spend, confirm your backup product supports the new platform and trial one low-priority VM first. The remainder then follows, one per maintenance slot.
List your current servers and their jobs. We reply with an outline configuration and a rough licence cost.
Your enquiry has reached us
You will hear back within one working day, and if you have reported an outage that is holding up work, it goes to the front of the queue.
No match for that name. Try a different spelling or pick a bigger town nearby - all our support is delivered online, so your choice has no effect on the service.