Sinkronisasi Basis Data Relasional menggunakan Paradigma Publish-Subscribe

Rifatullah, Muhammad (2018) Sinkronisasi Basis Data Relasional menggunakan Paradigma Publish-Subscribe. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
5114100118-Undergraduate_Theses.pdf - Published Version
Restricted to Repository staff only

Download (2MB) | Request a copy

Abstract

ITS mempunyai banyak sistem informasi. Namun, setiap sistem informasi mempunyai basis data masing-masing yang sebenarnya saling beririsan. Saat ini, setiap perubahan data harus dilakukan secara manual. Dengan jumlah data yang banyak dan perubahan yang masif, perubahan secara manual menjadi tidak efektif. Oleh sebab itu, dibutuhkan suatu sistem untuk sinkronisasi basis data. Pada tugas akhir ini, diajukan suatu metode untuk sinkronisasi basis data dengan pendekatan primary copy replication dan timestamp-based database synchronization dengan menggunakan paradigma publish-subscribe. Publish-subscribe digunakan untuk mengirimkan data dari tabel referensi ke tabel replika. Sinkronisasi diurutkan berdasarkan prioritas. Prioritas ditentukan oleh ketergantungan tabel tersebut ke tabel lain (dependensi). Tabel yang tidak mempunyai dependensi akan diproses terlebih dahulu. Uji coba pada tugas akhir ini menggunakan data jabatan (stakeholder) dan unitnya. Jumlah data stakeholder adalah 100 data, sedangkan unit 184 data. Skenario uji coba terdiri dari beberapa kondisi seperti tidak mempunyai dependensi, dependensi terhadap diri sendiri (self-reference), dan dependensi terhadap tabel lain pada satu basis data. Dari hasil uji coba, perubahan data pada tabel referensi bisa terdeteksi dan data berhasil dikirimkan ke tabel replika. Setelah data diproses oleh subscriber, data pada tabel referensi sama dengan tabel replika. Dengan menggunakan fitur RabbitMQ yaitu publisher confirms and message acknowledgements, data dapat dijamin sampai pada subscriber, walaupun broker atau subscriber mati. Hasil uji coba menunjukkan bahwa metode yang diajukan mampu melakukan sinkronisasi basis data dengan benar walaupun tabel mempunyai dependensi atau terjadi gangguan pada koneksi agar konsistensi dan integritas data dapat terjaga. ============= ITS has many information systems. However, every system has its own database which often overlaps with other database. Until now, every data change must be commited manually. With the number of data and massive changes, manual commit is not effective. Therefore, a database synchronization system is needed. This research proposes a method for synchronizing database using primary copy replication and timestamp-based database synchronization with publish-subscribe paradigm. Publish-subscribe is used to transfer data from reference table to its replicas. Synchronization is ordered by priority. Priority is determined by table’s dependency to another table. Table with no dependency will be processed first. The testing in this research uses data of ITS stakeholders and units. There are 100 records for stakeholders data, and 184 records for unit data. The testing has different conditions for each scenario such as dependency free, self-referential relation and table-to-table relation. The testing results show that data change can be detected and sent to replica successfully. After data is processed by subscriber, the data on reference table is equal to the data on replica table. With the help of the RabbitMQ capability of publisher confirm and message acknowledgements, the data can be guaranteed to consume by the subscriber, even though the broker or subscriber dies. The testing concluded that method proposed is capable of synchronizing table correctly even though it has dependency or network disruption while keeping consistency and data integrity.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Database; Synchronization; Publish/Subscribe Pattern; Primary Copy Replication; Timestamp Based Database Synchronization.
Subjects: Q Science > QA Mathematics > QA76 Computer software
Q Science > QA Mathematics > QA76.9 Computer algorithms.
Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4450 Databases
Divisions: Faculty of Information Technology > Informatics Engineering > (S1) Undergraduate Theses
Depositing User: Rifatullah Muhammad
Date Deposited: 24 Jan 2018 05:00
Last Modified: 24 Jan 2018 05:00
URI: http://repository.its.ac.id/id/eprint/50240

Actions (login required)

View Item View Item