|
How can the ‘InterCloud’ concept bring a revolution in computing?Bog'liq Cloud Computing dardingni sev, 2-topshiriq. O'zbek tili, 1-mavzu Zamonaviy AKT ning tushunchasi va tasnifi, expedia hisoboti, natija - 2022-10-28T091429.322, yangi MULOQOT psixologiyasi, AKT Sillabus KI(Kunduzgi) (1), Я хотел, Dropbox (1), Korporativ boshqaruv prinsiplar siyosat va amaliyot O‘rmonov 2019, 1111-dars, AXBOROT TEXNOLOGIYALARI HAQIDA TUSHUNCHA (2), Ekonometrik II amaliy (2), dsfdsg, dfvdfHow can the ‘InterCloud’ concept bring a revolution in computing?
Cloud computing has brought revolution but it is yet to reach its goal. In present-day scenario, multiple
vendors are offering cloud services being built-in either as proprietary or based on open technology.
But those services are not connected with each other. InterCloud or cloud federation concept talks
about establishing connection and communication among different cloud services from different
providers. The idea is adopted from networking system where network of internetworks built and
owned by different vendors eventually build the world wide internet system.
How is the disaster recovery planning associated with recovery point objective?
Keeping apart the disasters, even a small network interruption can cause loss of data on transition. Or
storage failure can cause data loss if back-up is not ready. Hence, data becomes safe only when it is ready
with well-planned backups. But backup is not created instantly with generation and primary storage of
new data. Backup is taken periodically. What transpires if any disaster happens at the middle of such a
period? The data generated since the last back-up point gets lost. There is no way of recovering this data.
System designers must take a strategic decision on how much of such data can be compromised
in case of disasters. The volume of such data is measured in time. For example, designers of some
system may decide that it is permissible to loss one business day’s data at the most. In such cases,
periodic backup must happen at each twenty-four hours’ interval. This fact is known as recovery
point objective (RPO) in disaster-recovery planning.
Why low-latency network communication is required for High-Performance Computing?
High-Performance Computing (HPC) facility is built over distributed computing environments.
Requirement of such system is to generate very high computing power for a relatively shorter
period of time. During such cases of high-performance computing, tasks execute in parallel at
distributed computing nodes and maintain direct communication with each other. This needs a
high-speed network communication path with very low latency so that computation does not get
disrupted due to slow network communication.
|
| |
