Lecture attendance data at universities is a reference in showing the credibility of each student used by lecturers as data for student grades as well as an evaluation material for the success of teaching and learning activities in lectures, but there are several examples of cases related to student attendance data currently prevalent in the world of education or lectures is the phenomenon of "Leave Absence" or better known as TA. In addition, other problems also arise from lecturers and administrative staff, namely difficulties in monitoring student attendance and efforts to validate attendance data because of the large amount of student data. Therefore in this study a system was proposed to reduce the level of fraud in filling the attendance list and effectiveness of student data processing using a system of applying the concept of the Internet of Things (IoT) with the fingerprint presence method. Existing system modeling results are expected to be able to support the service of processing academic data automatically and produce accurate and accurate statistical data and be able to reduce data manipulation factors from irresponsible parties.

The lecture data on college lectures into a reference in demonstrating the credibility of each student used by the lecturers as data for the student's value as well as the evaluation of the success of learning activities Teaching in the lecture, but there are some examples of cases, associated with the data of the student's presence that is currently involved in education or lectures is the phenomenon of "absent point". In addition, other problems also arise from the lecturers and administration officers, the difficulties in monitoring student attendance and efforts to validate the presences data because of the number of student data is so much. Therefore in this study submitted a system to reduce the level of fraud in filling the list of the presences and effectiveness of data processing students by using the system implementation of Face Recognition based on Open CV method with The Haar Cascade Classifier and Local Binary Patterns Histograms (LBPH) methods. The results of this Face Recognition study successfully detected when all the users that were reidentified were registered to the system, with the optimal range of Face Recognition to be detected to 150 cm. While Face Recognition is unsuccessful Detected when there is an obstacle covering the face objects and distances exceeding from 150 cm.

Parking areas in urban centers are increasingly limited, but not in line with technological developments in addressing the problem of parking service needs in the urban public service center by implementing smart city technology. Many times the parking service waiting time is too long, due to a complicated procedure. Waiting time for services that are too long makes the parking queue increase in length, especially when registering for the end of a new parking lot. Besides, the storage of visitors’ parking slip accumulation is still manual and not paperless. Therefore, in this study, a solution was proposed in the form of a system to accelerate administrative services for parking visitors by integrating Radio Frequency Identification (RFID) and Wireless Sensor Network (WSN) technologies. This system utilizes the database so that the results of data processing can be used by management to obtain visitor data information on parking online and quickly. The results obtained from this study are a prototype of a parking visitor monitoring system using WSN technology based on Zigbee and RFID protocols. Meanwhile, the results of visitor parking information presented in data processing can be used by management to find out statistical data relating to parking services.

No abstract

Seiring perkembangan teknologi yang begitu pesat, telah muncul banyak metode untuk manajemen dan analisis log dari sebuah komputer diantaranya metode Grafana Loki dan ELK Stack. Sehingga dampak dari perkembangan ini menimbulkan banyak variasi dan ketidaktahuan para administrator dalam menentukan metode mana yang sesuai dengan kebutuhan mereka. Pada penelitian ini menganalisis performa dari kedua metode tersebut terhadap server honeypot saat terjadi serangan dengan parameter penggunaan CPU dan Memori, kedua parameter tersebut merupakan standar untuk para administrator dalam mempertimbangkan metode yang akan dipilih. Kesimpulan dari penelitian ini bahwa berdasarkan parameter yang digunakan metode Grafana Loki lebih efisien dari segi pemakaian CPU dan Memori dibandingkan metode ELK Stack, Grafana Loki sangat ringan untuk diimplementasikan tetapi dengan fitur yang terbatas, sedangkan ELK Stack lebih banyak memakai resource CPU dan Memori tetapi mempunyai fitur yang lebih lengkap.Kata Kunci : Performa, Honeypot, ELK Stack, Grafana Loki