I’m a Brit and, hands up, we are obsessed with the weather. Mind you September was one of driest on record, in fact the driest since 1910 contrasted by October, which is looking very wet from where I’m sitting. I say obsessed as I have to log an interest having spent a proportion of my career launching weather balloons, taking temperatures, making observations and drawing charts. As for charts here’s one for you. The most complex graph you are ever likely to see. It’s called a Tephigram. And thankfully children do not get to complete these for their homework. It’s certainly degree level work!
They show the vertical state of the atmosphere. Great for pilots who need to plan their route to avoid headwinds and weather aloft. You would not believe the aircraft fuel savings you can achieve by understanding this stuff.
But look at the observational data that comes from the balloons.
TTAA 51231 03808 99996 07819 17005 00057 ///// ///// 92698 03843 20018 85379 00356 20020 70934 04976 22032 50550 19143
27537 40713 29940 27552 30912 44335 27069 25032 54327 28571 20171 61931 29570 15353 54782 27047 10614 54982 25523
88207 62929 29586 77214 29090 42341 31313 48008 82315
Anyone that tells you that compression and deduplication give you consistent efficiency savings for all data types and workloads is talking bull. And as I have said before, and will say again, Caveat Emptor! You mileage will, and does, vary. Understand your workload. Do your homework. Talk to NetApp.
I was pleased to see a story this week from the German Meteorological Service, Deutscher Wetterdienst, DWD. Weather forecasting and climate modelling are one of the most data intensive applications I have come across. On a par with oil and gas exploration, and certain nuclear modelling applications, which shall remain nameless. After the initial observations are collected and quality controlled they form the basis for forecasting the weather hours, days, or weeks in advance, or in the case of the climate, maybe 100 years into the future. In weather forecasting time is critical as products need to be distributed to waiting customers, such as airlines, emergency services and commercial customers. DWD had identified a requirement for high performance storage to accelerate the forecast process and deliver savings. They selected the NetApp E5500 E-Series to take part in a competitive head to head. They found that the E5500 surpassed their expectation by meeting the 30GB per second throughput requirement while also requiring 48% less floor space, and reduced energy consumption by 30%. Needless to say the E5500 was selected. Personally I have seen E-Series selected as the storage of choice in High Performance Computing environments many times. I remember walking into one oil and gas exploration company and their online storage was wall to wall E-Series. I was therefore interesting to read the confirmation of what I had already experienced by Dr. Henning Weber, Head of System Support in the Central Systems department of the DWD.
“The NetApp E-Series is compact, modern, and extendable. The technology has a very high level of maturity and has so far proven to be extremely reliable. The solution also uses little energy and space. These characteristics make the systems ideal for the specific requirements of the DWD in the field of high-performance computing”
If you are looking at High Performance Storage then I urge you to evaluate the E-Series solutions. Other sweet spot workloads include Media, Broadcast and Finance. You will find the DWD story here:
Follow me on Twitter @lozdjames