Compute sized to the workload, not to the quote you were given
Dell EMC PowerEdge and HPE ProLiant rack, tower and hyperconverged compute — specified against measured load, supplied at partner pricing, racked, configured and supported by the same engineers.
Servers, end to end
Specification
CPU generation and core count, memory channels populated properly, disk and controller chosen for the actual IO pattern. Most quoted servers are wrong in at least one of these.
Supply
Dell EMC PowerEdge and HPE ProLiant at partner pricing, with the warranty tier and response time that matches how critical the box is.
Rack and build
Racking, cabling, power distribution, labelling and firmware brought to a supported baseline before anything goes live.
Hyperconverged
Where it earns its place, a hyperconverged cluster instead of separate compute and storage. Where it does not, we say so.
Out-of-band management
iDRAC and iLO configured, on a separate management network, with alerting that reaches a person. This is the difference between a failed disk and an outage.
Lifecycle and support
Warranty tracked, firmware on supported branches, capacity reviewed, and end of life planned rather than discovered.
Measured, not guessed
Almost every underperforming server we inherit was specified from a template.
1. Measure
CPU, memory, disk latency and queue depth from the existing estate. If there is nothing to measure, we model against the application vendor’s real requirements.
2. Size
Headroom for growth and for a node loss, not just for today. A cluster sized with no failure margin is a single point of failure with extra steps.
3. Quote with reasoning
Every line item has a reason attached. You should be able to challenge any of them.
4. Build and hand over
Racked, configured, documented and handed over with an as-built record.
Common questions
Dell or HPE?
We are a partner for both, so we have no reason to prefer one. The decision usually comes down to what your team already operates, existing support contracts, and lead time in Egypt at the moment you order.
Can you support servers we already have?
Yes. We start by documenting what is installed, its warranty position and its firmware level, which for most clients is the first time that has been written down.
Do we need new hardware at all?
Sometimes not. We have told clients to add memory and move on. If the honest answer is that your existing estate has years left, you will hear that.
What about the rest of the stack?
Servers sit under storage, virtualisation and backup. We design them together because they fail together.
Send us what you run today.
A server inventory and a note of what is slow is enough for us to tell you whether you have a sizing problem or a configuration problem.
02 3537 5791