It is a fact of life that, over time, the documentation for servers’ room and data centres becomes less and less accurate and up-to-date. Often, the cabling becomes less than perfectly tidy too.
In real life, not every move, add or change gets properly recorded. Turnover of staff and the effects of emergency fault-finding just make it worse.
As a result, fault rectification becomes more difficult resulting in extended downtime.
Every so often, it is good practice to do a full audit and establish a new, fully accurate, documentation baseline. This simplifies ongoing management and makes fault-finding significantly faster reducing your downtime and business interruption.
To ensure that the servers can be moved, re-installed and functioning in the least possible downtime, the new location needs to be successfully pre-wired and ready to receive the equipment.
This requires that an accurate map of each of the servers, switch and ancillary equipment documented – together with their exact interconnections – must be prepared in advance of the move.
All of this documentation will be in place following an IT audit.
You will receive documentation which maps all equipment and locations, together with a record of all connected physical ports, both management and network.
In addition to the standard audit, Connectium will physically trace every connection in your datacentre, creating a complete connectivity map detailing how the servers are interconnected and documenting all critical information.
No ports are disconnected during this physical trace which is undertaken by skilled engineers.
For higher-risk environments, we tailor our service according to your needs. Often, our clients opt for an out of hours physical trace.
To ensure that your data centre documentation remains accurate after your audit, Connectium can devise an effective bespoke solution for your people to operate day-to-day. We can also add quality control using a programme of regular re-audits if you desire.
The Connectium team has been conducting datacentre audits (and large data centre migrations) for almost a decade.
In that time, hundreds of clients have enjoyed this service to make their documentation 100% accurate and up-to-date.
Connectium audits are virtually risk-free because we only use non-invasive techniques; meaning there is no need for any disconnection or downtime.
Don’t take risks with your data centre, call on the safest hands in the industry to put you back in control.
The power, cooling and physical space that underpins all modern IT systems are often overlooked. But there is plenty of innovation in this hidden world that supports all the virtual ones.
Software
Perhaps the most surprising aspect of critical infrastructure is how much innovation is being driven by software.
The control systems for power, cooling and even security, are all software systems. The same trends we see in the broader IT market can be seen in critical infrastructure.
Major suppliers are building software platforms with APIs that customers, both data centres and end-customers themselves, can plug into to find more efficient ways to consume resources and just keep tabs on their equipment housed within the facility.
“The biggest gains are efficiencies driven by software,” says John Atherton, general manager, power quality for Eaton Australia and New Zealand. The vendor’s Intelligent Power Manager software has been certified as VMware-ready and integrates with VMware’s vRealize Operations Manager software, for example.
Mark Deguara, director of data centre solutions at Emerson Network Power Australia, says adding monitoring equipment is the first step to making even legacy data centres more efficient. Emerson’s Trellis platform for thermal management offers real-time monitoring of heat in a data centre, so providers can adjust systems to respond in real-time.
“The data centre becomes autonomous,” says Deguara. “Optimisation of legacy data centres starts with getting metering in and using it. First your measure, then your baseline and then you improve.”
It’s all part of a movement towards virtual data centres, where individual data centres work together like blades in a blade-chassis and workloads are shared between them. Coordinating the workloads and all of the critical infrastructures they rely on requires the new sophisticated software platforms that run modern data centres.
Power
Without electrons there would be no computing, so keeping all the IT equipment fed with a steady supply of clean and healthy power is perhaps the most critical part of all critical infrastructure. The biggest recent development in power is in energy storage, in batteries. Storing energy on-site is required for uninterruptible power supply (UPS) if the main power feed fails, but how that power is stored can vary enormously.
Most UPS units use banks of lead-acid batteries to store enough power to run the data centre for a short period if the grid power feed fails, but newer battery technologies with superior features are becoming more affordable.
Eaton has entered into a partnership in Europe with carmaker Nissan to reuse batteries from the Nissan Leaf electric car. The lithium-ion batteries used to store power for electric and hybrid vehicles stop being viable when they drop to about 85% efficiency, so they’re no longer useful for cars, but adding Eaton’s power management software to the systems can give them new life as data centre power storage.
Lithium-ion battery systems cost about 1.5 to 3 times that of the standard lead-acid batteries UPS systems have used for decades. By reusing batteries from Nissan, Eaton can tap into a supply that is substantially cheaper than newly made ones, making the economics more compelling, given the superior features regarding maintenance.
Since data centres need to store energy locally anyway, renewable generation sources such as wind or solar can be used to supplement grid feeds to help bring costs down. While renewables can’t yet provide the bulk of a data centre’s substantial power needs, they can help to improve efficiency, which reduces costs.
“The marketing benefits of going green are also becoming an important aspect of data centre providers’ pitch to corporate customers looking for
lower TCO.
Cooling
Keeping all the equipment cool is the second-most important part of critical infrastructure. Cooling without major additional cost is a key part of keeping costs down.
“We use mostly free-air and evaporative cooling,” says Josh Griggs, managing director of data centre provider Metronode. “We have sensors throughout the facility automatically adjusting the cooling mix as needed. We can have a 1kW rack running right next to a 30kW rack with no problems.”
Free-air cooling mixes some outside air with data centre air when the external air is cooler. The air is filtered to remove pollutants like smoke or dust and, when clean and cool, can save a lot of energy.
With evaporative cooling, warm air is cooled by evaporating water, using the same principle as the old Coolgardie Safe bush fridge. Modern systems don’t mix the cooled air with the water itself, to keep humidity within the data centre low.
Inside the data centre itself, modern designs put the cool air underneath the floor, since hot air rises. Rather than cooling the entire room, only the racks themselves get cooled through channelling equipment. Carefully monitoring the temperature throughout the data centre and adjusting airflow as needed ensures the right amount of cooling can be guided to where it is required.
Right at the edge of cooling technology, liquid cooling is starting to make a comeback.
“Market acceptance of liquid cooling isn’t there yet,” says Griggs, “but it’s only a few years of R&D away.”
Dell EMC recently released its Triton liquid-cooled server technology, developed in partnership with customer eBay. As this kind of technology filters down from the very large-scale operators it will start to be offered by data centre providers as an option. We can expect liquid cooling to supplement traditional air cooling in just a few years as companies seek ever denser compute.
Security
Physical security is one of those things that we take for granted, but here things are changing as well. Data centres that host multiple tenants need to protect each tenant from the other, as well as outside threats. “Security and uptime are table-stakes,” says Griggs. While every data centre CRN has visited recently has the same array of security guards, cameras everywhere and multi-stage locks, electronic access tracking is becoming more granular.
Some facilities, including Equinix’s latest SY4 facility in Sydney, offer swipe-card access for individual racks instead of physical keys. The big advantage of swipe cards over keys is that swipe-card access provides a solid audit trail of who accessed which rack and when. An audit trail can be very useful when trying to figure out who mis-cabled the server, not just for review after a security incident.
Remote sites
Another aspect of critical systems that’s easy to ignore is that there are lots and lots of IT systems that don’t live in a data centre. Think of all the mobile base-stations out there providing LTE and better data services to the army of mobile devices. Modern farms are full of electronics – sensors, actuators, cameras, you name it – and they need critical infrastructure.
Rural areas are hit with sun, dust, wind and other elements that quickly destroy standard data centre gear. To survive in these conditions requires specifically designed equipment.
“We ran a project with battery maker Redflow at sites that used to use on-site generators for power,” says Deguara. The project combined on-site renewable energy generation using solar and wind power, with energy storage, using Redflow batteries.
“A site that would have required a generator to run 24 hours a day could use a generator for just 3.5 hours a day,” says Deguara.
Apart from the environmental savings, there are immense cost savings from reduced fuel consumption, less generator maintenance and humans not needing to attend remote sites.
“It’s all about the economics,” says Deguara. “If the marketplace isn’t willing to take on the technology, it’s not important.” “
Source: https://www.crn.com.au/feature/inside-tomorrows-smarter-data-centre-441061
Tags: centre, audit, management, operations, condition, maintenance, experience, one, site, fire, new data, different, provide, identify, industry, could, objectives, centre, audit, data centre design, designed, place, facility, focus, financial, work, standards, environment, operation, centre, audit, tier, information, review, case, points, please, service, group, future, know, approach, recommendations, comprehensive, team, process, also, centre, audit, professional, detailed, current, services, failure, report, analysis, requirements, assessment, data centre audit, plan, infrastructure, operational, data, business, risk, standards, environment, operation, centre, audit, tier, information, review, case, points, please, service, group, future, know, approach, recommendations, comprehensive, centre, audit, team, process, also, professional, detailed, current, services, failure, report, analysis, requirements, assessment, data centre audit, centre, audit, plan, infrastructure, operational, data, business, risk, recommendations, centre, audit, comprehensive, team, process, also, professional, detailed, current, services, failure, report, analysis, requirements, assessment, centre, audit, data centre audit, plan, infrastructure, operational, data, business, risk, privacy, uses, ensure, data centre audits, european, within, centre, audit, including, improvements, access control, support, report, analysis, requirements, centre, audit, assessment, data centre audit, plan, infrastructure, operational, data, business, risk, available, email, best, cooling systems, findings, centre, audit, existing data, list, require, facilities, organisation, processes, plan, infrastructure, operational, data, business, risk, availability, centre, audit, centre, audit, aspects, many, procedures, key, mechanical, risks, office, management, operations, condition, maintenance, experience, one, site, infrastructure, operational, data, business, risk, fire, new data, different, provide, identify, industry, could, objectives, data centre design, designed, place, facility, business, risk, focus, financial, work, standards, environment, operation, tier, information, review, case, points, please, service, group, future, know, approach, recommendations, comprehensive, team, process, also, professional, detailed, current, services, failure, report, analysis, requirements, assessment, data centre audit, plan, infrastructure, operational, data, business, risk, view, energy efficiency, data centre audits, document, audit data, electrical systems, European, reduce, compliance, relevant, floor, environmental, set, consent, access, data centres, email, including, issues, include, cooling systems, aspects, findings, required, level, improve, business continuity, existing data, access control, list, require, support, facilities, organisation, processes, levels, type, procedures, availability, thermal, training, operations, operating, identify, cost, specialist, objectives, data centre design, mechanical, key, focus, financial, costs, condition, process, critical, services, cooling, security, review, power, energy, electrical, provide, recommendations, facility, report, environment, design, requirements, standards, efficiency, comprehensive, failure, analysis, future, understanding, approach, assessment, maintenance, detailed, current, operation, risk, operational, plan, business, infrastructure, systems, centre, team, operational, condition, process, management, also, business, plan, maintenance, detailed, current, need, data centre audit, risk, services, within, information, email, best, require, facilities, contact, levels, key, training, service, identify, mechanical, infrastructure, report, risks, recommendations, requirements, would, standards, operational, data