State-of-the-Art ICT Infrastructure Supports Morocco's Online Social Security System
The National Social Security Fund of Morocco — formally, la Caisse Nationale de Sécurité Sociale (CNSS) — is the public institution responsible for managing and making social security payments, from health insurance to retirement and disability benefits. As such, it collects social security contributions from both employers and employees, while also always looking for ways to improve the overall system, refining policies and programs that address the ever-evolving needs of the population at large.
An agency working under the Ministry of Economy and Finance, CNSS has a total of 13 regional offices that are responsible for day-to-day work. This includes a relatively new Online Transaction Processing (OLTP) system, which is tasked with handling a large volume of traffic. Keeping this data secure yet available to staff is, of course, key to efficient operations. And with every employer operating in the country also interacting with CNSS — each must register every employee, filing their salaries month-to-month — the OLTP system must have rapid, real-time response times or risk leaving end-users dissatisfied. All of this requires high performing Information and Communications Technology (ICT) infrastructure.
The Huawei solution chosen includes two sets of OceanStor Dorado Non-Volatile Memory express (NVMe) storage to host core business applications. Forming an active-active storage array and DR solution, this new system delivers the high availability and performance required for CNSS's core services.
Huawei's FlashLink algorithms are one of the major features of the OceanStor Dorado series, ensuring high performance, through a multi-core algorithm that enables the separation of different workload groups: all groups work separately without interacting, boosting computing power. This is a great fit for CNSS's OLTP system, which has to handle multiple applications at the same time, ensuring uninterrupted operations even during the peak period.
The storage solution features an Artificial Intelligence (AI) chip and intelligent cache algorithm that actively analyzes and understands Input/Output (I/O) rules based on Machine Learning (ML) frameworks, increasing CNSS's cache hit ratio by more than 50% compared to the legacy infrastructure. This brings latency down and makes the overall online user-experience much smoother.
Multi-streaming technology also separates metadata into different segments, such as hot and cold data, based on the frequency of modifications, increasing trash collection efficiency since all hot data is recycled immediately, without moving cold data at all. As a result, when the OLTP system modifies small amounts of data, read and write operations remain balanced, maximizing system efficiency for more important tasks.