ANALISIS UTILISASI RESOURCE CLUSTERS PADA HADOOP MENGGUNAKAN VIRTUALIZATION

I Kadek Susila Satwika (Institut Bisnis dan Teknologi Indonesia)
I Putu Susila Handika (Institut Bisnis dan Teknologi Indonesia)
Made Hanindia Prami Swari (UPN “Veteran” Jawa Timur)

Jurnal RESISTOR (Rekayasa Sistem Komputer)
Vol. 5 No. 1 (2022): Jurnal RESISTOR Edisi April 2022

Large amounts of data processing necessitate the use of a dependable infrastructure. Using clustering technology in Big Data processing is a solution for faster processing times. This study evaluated the performance of the Hadoop server using virtualization technology. A varying number of query requests are sent to Hadoop server clusters at the same time. Then, the CPU and memory (RAM) utilization was calculated. According to the test results, the CPU usage on the namenode reached 100% at the start of the process, followed by an increase in CPU usage on the datanode the next time. Meanwhile, the namenode uses the most memory when it receives 25 requests at once. This demonstrates that the namenode can only serve a maximum of 25 requests at the same time.