Analisis Komparatif Performa Cloud Open Source Harvester, Cockpit Cloud, Dan LXD Pada Arsitektur Cloud Badan Informasi Geospasial

Sihombing, Andreas Timotius Parhorasan (2025) Analisis Komparatif Performa Cloud Open Source Harvester, Cockpit Cloud, Dan LXD Pada Arsitektur Cloud Badan Informasi Geospasial. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5027211019-Undergraduate_Thesis.pdf] Text
5027211019-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (19MB) | Request a copy

Abstract

Layanan cloud di lingkungan BIG dijalankan menggunakan platform virtualisasi komersial VMware sebagai solusi utama. Namun, seiring adanya kebijakan efisiensi anggaran, diperlukan alternatif berbasis open source yang lebih ekonomis dan mandiri. Penelitian ini bertujuan untuk melakukan analisis komparatif terhadap tiga platform manajemen cloud open source, yaitu Harvester, Cockpit Cloud, dan LXD dalam mendukung layanan private cloud BIG, khususnya sistem Palapa. Evaluasi dilakukan melalui implementasi aplikasi cloud internal BIG serta pengujian performa teknis yang meliputi komputasi (CPU, memori, disk I/O), jaringan internal dan eksternal, serta ketahanan terhadap beban aplikasi. Proses pengujian menggunakan Sysbench, iperf3, dan JMeter. Hasil pengujian menunjukkan bahwa masing-masing platform cloud memiliki keunggulan yang berbeda. LXD memiliki performa komputasi terbaik dengan rata-rata kecepatan CPU single-thread sebesar 15.884 event/s dan multi-thread sebesar 46.155 event/s, serta kecepatan disk I/O read dan write sebesar 427 MiB/s. Sementara itu, Cockpit Cloud memiliki performa memori terbaik dengan kecepatan read 32.989 MiB/s dan write 47.336 MiB/s pada single VM. Selain itu, Cockpit Cloud juga menunjukkan ketahanan paling stabil terhadap peningkatan beban pengguna hingga 2.000 user, dengan error rate maksimum sebesar 7,27% dan response time tertinggi 6.895 ms. Platform Harvester menunjukkan performa yang seimbang, meskipun memiliki kelemahan pada aspek jaringan eksternal dengan packet loss sebesar 1,92 persen.
========================================================================================================================
Cloud services within BIG are operated using the commercial virtualization platform VMware as the primary solution. However, with the introduction of budget efficiency policies, there is a need for a more economical and independent open source-based alternative. This study aims to conduct a comparative analysis of three cloud open source management platforms, namely Harvester, Cockpit Cloud, and LXD, to support BIG's private cloud services, particularly for the Palapa system. The evaluation was carried out through the implementation of BIG's internal cloud application and technical performance testing covering computing (CPU, memory, disk I/O), internal and external network, as well as application load resilience. The testing process utilized Sysbench, iperf3, and JMeter.
The results show that each cloud platform has its own strengths. LXD demonstrated the best computing performance with average CPU single-thread speed of 15,884 event/s and multi-thread speed of 46,155 event/s, as well as disk I/O read and write speeds of 427 MiB/s. Meanwhile, Cockpit Cloud achieved the best memory performance with read speed of 32,989 MiB/s and write speed of 47,336 MiB/s on a single VM. Cockpit Cloud also showed the most stable resilience under increasing user load up to 2,000 users, with a maximum error rate of 7.27\% and peak response time of 6,895 ms. The Harvester platform exhibited balanced performance overall, although it had a drawback in external network with a packet loss rate of 1.92 percent.

Item Type: Thesis (Other)
Uncontrolled Keywords: Cloud Computing, Harvester, Cockpit Cloud, LXD, Benchmarking, Cloud Architecture, Geospatial. Cloud Computing, Harvester, Cockpit Cloud, LXD, Benchmarking, Arsitektur Cloud, Geospasial.
Subjects: Q Science > QA Mathematics > QA76.585 Cloud computing. Mobile computing.
Divisions: Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis
Depositing User: Andreas Timotius Parhorasan Sihombing
Date Deposited: 18 Jul 2025 06:52
Last Modified: 18 Jul 2025 06:52
URI: http://repository.its.ac.id/id/eprint/120020

Actions (login required)

View Item View Item